Version: 3.7
weeWX Customization Guide

This document covers the customization of weeWX. It assumes that you have read, and are reasonably familiar with, the Users Guide.

Introduction

This document starts with an overview of the architecture of weewx. If you are only interested in customizing the generated reports you can probably skip the overview and proceed directly to the section The Standard skin.conf . With this approach you can easily add new plot images, change the titles of images, change the units used in the reports, and so on.

However, if your goal is a specialized application, such as adding alarms, RSS feeds, etc., then it would be worth your while to read about the internal architecture.

Most of the guide applies to any hardware, but the exact data types are hardware-specific. Many examples use types specific to the Davis Vantage series. Unless you are using an unusual type you are unlikely to run into trouble.

Warning!
weeWX is still an experimental system and, as such, its internal design is subject to change. Future upgrades may break any customizations you have done, particularly if they involve the API (skin customizations tend to be more stable).

Where to put customizations

For configuration changes, simply modify the weeWX configuration file weewx.conf. Customization of reports may require changes to a skin configuration file skin.conf or template files ending in .tmpl. The weewx.conf configuration file and the skin.conf configuration and template files in the skins directory will be preserved when you upgrade.

Other customizations require new Python code or modifications of example code. Where should you put the code? If you simply modify the examples in place, then your changes will be overwritten the next time you do an upgrade.

A better idea is to put the code in the user, directory. For example, copy example code from the examples directory to the user directory, then modify it there. If your modification does not contain much code, consider putting it in the file extensions.py in the user directory. Because the user directory is preserved through upgrades, you won't have to redo any changes you might have made.

Data architecture

WeeWX is data-driven. When the sensors spit out some data, weeWX does something. The 'something' might be to print out the data, or to generate an HTML report, or to use FTP to copy a report to a web server, or to perform some calculations using the data. These actions are called services.

A driver is Python code that communicates with the hardware. The driver reads data from a serial port or a device on the USB or a network interface. It handles any decoding of raw bits and bytes, and puts the resulting data into LOOP packets and archive RECORDs. Whenever possible, weeWX should work "out of the box" with minimal configuration. However, in some cases the driver may include code to configure and test the hardware, typically via the wee_device utility.

In many cases there are useful dependent variables that use the primary variables emitted by the driver. For example, rain rate, windchill, heatindex, humidex, apptemp, ET are all dependent quantities — they require primary observations such as wind speed, temperature, solar radiation. The firmware for some hardware calculates dependent quantities. The weeWX service StdWXCalculate fills in the gaps and determines whether to use a value from firmware or a calculation from weeWX. Sometimes the firmware simply does it wrong.

LOOP packets vs. archive records

Generally, there are two types of data that flow through weeWX: LOOP packets, and archive records.

LOOP packets

LOOP packets are the raw data generated by the device driver. They get their name from the Davis Instruments documentation. For some devices they are generated at rigid intervals, such as every 2 seconds for the Davis Vantage series, for others, irregularly, every 20 or 30 seconds or so. LOOP packets may or may not contain all the data types. For example, a packet may contain only temperature data, another only barometric data, etc. These kinds of packet are called partial record packets. By contrast, other types of hardware (notably the Vantage series), every LOOP packet contains every data type.

In summary, LOOP packets can be highly irregular, but they come in frequently.

Archive records

By contrast, archive records are highly regular. They are generated at regular intervals (generally every 5 to 30 minutes), and all contain the same data types. They represent an aggregation of the LOOP packets over the archive interval. The exact kind of aggregation depends on the data type. For example, for temperature, it's generally the average temperature over the interval. For rain, it's the sum of rain over the interval. For battery status it's the last value in the interval.

Some hardware is capable of generating its own archive records (the Davis Vantage and Oregon Scientific WMR200, for example), but for hardware that does not archive data, weeWX generates them.

It is the archive data that is put in the SQL database, although, occasionally, the LOOP packets can be useful (such as for the Weather Underground's "Rapidfire" mode).

The weeWX service architecture

At a high-level, weeWX consists of an engine class called StdEngine. It is responsible for loading services, then arranging for them to be called when key events occur, such as the arrival of LOOP or archive data. The default install of weeWX includes the following services:

The standard weeWX services
Service Function
weewx.engine.StdTimeSynch Arrange to have the clock on the station synchronized at regular intervals.
weewx.engine.StdConvert Converts the units of the input to a target unit system (such as US or Metric).
weewx.engine.StdCalibrate Adjust new LOOP and archive packets using calibration expressions.
weewx.engine.StdQC Check quality of incoming data, making sure values fall within a specified range.
weewx.wxservices.StdWXCalculate Calculate any missing, derived weather observation types, such a dewpoint, windchill, or altimeter-corrected pressure.
weewx.engine.StdArchive Archive any new data to the SQL databases.
weewx.restx.StdStationRegistry
weewx.restx.StdWunderground
weewx.restx.StdPWSweather
weewx.restx.StdCWOP
weewx.restx.StdWOW
weewx.restx.StdAWEKAS
Various RESTful services (simple stateless client-server protocols), such as the Weather Underground, CWOP, etc. Each launches its own, independent thread, which manages the post.
weewx.engine.StdPrint Print out new LOOP and archive packets on the console.
weewx.engine.StdReport Launch a new thread to do report processing after a new archive record arrives. Reports do things such as generate HTML or CSV files, generate images, or FTP/rsync files to a web server.

It is easy to extend old services or to add new ones. The source distribution includes an example new service called MyAlarm, which sends an email when an arbitrary expression evaluates True. These advanced topics are covered later in the section Customizing the weeWX service engine.

The standard reporting service, StdReport

For the moment, let us focus on the last service, weewx.engine.StdReport, the standard service for creating reports. This will be what most users will want to customize, even if it means just changing a few options.

Reports

The standard reporting service, StdReport, runs zero or more reports. The specific reports which get run are set in the configuration file weewx.conf, in section [StdReport].

The default distribution of weeWX includes three reports:

Report Default functionality
StandardReport Generates day, week, month and year "to-date" summaries in HTML, as well as the plot images to go along with them. Also generates NOAA monthly and yearly summaries.
FTP Arranges to upload everything in the HTML_ROOT directory up to a remote webserver.
RSYNC Like FTP, but uses rsync for transferring files to a remote webserver.

Note that the FTP and RSYNC "reports" are a funny kind of report in that it they do not actually generate anything. Instead, they use the reporting service engine to arrange for things to be transferred to a remote server.

Skins

Each report has a skin associated with it. For most reports, the relationship with the skin is an obvious one: it contains the templates, any auxiliary files such as background GIFs or CSS style sheets, and a skin configuration file, skin.conf. If you will, the skin controls the look and feel of the report. Note that more than one report can use the same skin. For example, you might want to run a report that uses US Customary units, then run another report against the same skin, but using metric units and put the results in a different place. All this is possible by either overriding configuration options in the weeWX configuration file weewx.conf or the skin configuration file skin.conf.

Like all reports, the FTP and RSYNC "reports" also use a skin, and include a skin configuration file, although they are quite minimal.

Skins live in their own directory called skins, whose location is referred to as SKIN_ROOT.

Generators

To create their output, skins rely on one or more generators, which are what do the actual work, such as creating HTML files or plot images. Generators can also copy files around or FTP/rsync them to remote locations. The default install of weeWX includes the following generators:

Generator Function
weewx.cheetahgenerator.CheetahGenerator Generates files from templates, using the Cheetah template engine. Used to generate HTML and text files.
weewx.imagegenerator.ImageGenerator Generates graph plots.
weewx.reportengine.FtpGenerator Uploads data to a remote server using FTP.
weewx.reportengine.RsyncGenerator Uploads data to a remote server using rsync.
weewx.reportengine.CopyGenerator Copies files locally.

Note that the three generators FtpGenerator, RsyncGenerator, and CopyGenerator do not actually generate anything having to do with the presentation layer. Instead, they just move files around.

Which generators are to be run for a given skin is specified in the skin's configuration file skin.conf, in section [Generators].

Templates

A template is a text file that is processed by a template engine to create a new file. weeWX uses the Cheetah template engine. The generator weewx.cheetahgenerator.CheetahGenerator is responsible for running Cheetah at appropriate times.

A template may be used to generate HTML, XML, CSV, Javascript, or any other type of text file. A template typically contains variables that are replaced when creating the new file. Templates may also contain simple programming logic.

Each template file lives in the skin directory of the skin that uses it. By convention, a template file ends with the .tmpl extension.

The utility wee_reports

If you make changes, how do you know what the results will look like? You could just run weeWX and wait until the next reporting cycle kicks off but, depending on your archive interval, that could be a 30 minute wait or more.

The utility wee_reports allows you to run a report whenever you like. To use it, just run it from a command line, with the location of your configuration file weewx.conf as the first argument. Optionally, if you include a unix epoch timestamp as a second argument, then the report will use that as the "Current" time; otherwise, the time of the last record in the archive database will be used. Here is an example, using 1 May 2014 00:00 PDT as the "Current" time.

wee_reports weewx.conf 1398927600

For more information about wee_reports, see the Utilities Guide

The database

WeeWX uses a single database to store and retrieve the records it needs. It can be implemented by using either SQLITE3, an open-source, lightweight SQL database, or MySQL, an open-source, full-featured database server.

Structure

Inside this database are several tables. The most important is the archive table, a big flat table, holding one record for each archive interval, keyed by dateTime, the time at the end of the archive interval. It looks something like this:

Structure of the archive database table
dateTime usUnits interval barometer pressure altimeter inTemp outTemp ...
1413937800 1 5 29.938 null null 71.2 56.0 ...
1413938100 1 5 29.941 null null 71.2 55.9 ...
... ... ... ... ... ... ... ... ...

The first three columns are required. Here's what they mean:

Name Meaning
dateTime The time at the end of the archive interval in unix epoch time. This is the primary key in the database. It must be unique, and it cannot be null.
usUnits The unit system the record is in. It cannot be null. See the Appendix: Units for how these systems are encoded.
interval The length of the archive interval in minutes. It cannot be null.

In addition to the main archive table, there are a number of smaller tables inside the database, one for each observation type, that hold daily summaries of the type. For example, the minimum and maximum value seen during the day, and at what time. These tables have names such as archive_day_outTemp or archive_day_barometer. Their existence is generally transparent to the user.

Binding names

While most users will only need the one weather database that comes with WeeWX, the reporting engine allows you to use multiple databases in the same report. For example, if you have installed the cmon computer monitoring package, which uses its own database, you may want to include some statistics or graphs about your server in your reports, using that database.

An additional complication is that weeWX can use more than one database implementation: SQLite or MySQL. Making users specify in the templates not only which database to use, but also which implementation, would be unreasonable.

The solution, like so many other problems in computer science, is to introduce another level of indirection, a database binding. Rather than specify which database to use, you specify which binding to use. Bindings do not change with the database implementation, so, for example, you know that wx_binding will always point to the weather database, no matter if its implementation is a sqlite database or a MySQL database. Bindings are listed in section [DataBindings] in weewx.conf.

The standard weather database binding that weeWX uses is wx_binding. This is the binding that you will be using most of the time and, indeed, it is the default. You rarely have to specify it explicitly.

Programming interface

WeeWX includes a module called weedb that provides a single interface for many of the differences between database implementations such as SQLite and MySQL. However, it is not uncommon to make direct SQL queries within services or search list extensions. In such cases, the SQL should be generic so that it will work with every type of database.

The database manager class provides methods to create, open, and query a database. These are the canonical forms for obtaining a database manager.

If you are opening a database from within a weeWX service:

db_manager = self.engine.db_binder.get_manager(data_binding='name_of_binding', initialize=True)

# Sample query:
db_manager.getSql("SELECT SUM(rain) FROM %s "\
    "WHERE dateTime>? AND dateTime<=?" % db_manager.table_name, (start_ts, stop_ts))

If you are opening a database from within a weeWX search list extension, you will be passed in a function db_lookup() as a parameter, which can be used to bind to a database. By default, it returns a manager bound to wx_binding:

wx_manager    = db_lookup()                                    # Get default binding
other_manager = db_lookup(data_binding='some_other_binding')   # Get an explicit binding

# Sample queries:
wx_manager.getSql("SELECT SUM(rain) FROM %s "\
    "WHERE dateTime>? AND dateTime<=?" % wx_manager.table_name, (start_ts, stop_ts))
other_manager.getSql("SELECT SUM(power) FROM %s"\
    "WHERE dateTime>? AND dateTime<=?" % other_manager.table_name, (start_ts, stop_ts))

If opening a database from somewhere other than a service, and there is no DBBinder available:

db_manager = weewx.manager.open_manager_with_config(config_dict, data_binding='name_of_binding')

# Sample query:
db_manager.getSql("SELECT SUM(rain) FROM %s "\
    "WHERE dateTime>? AND dateTime<=?" % db_manager.table_name, (start_ts, stop_ts))

The DBBinder caches managers, and thus database connections. It cannot be shared between threads.

Units

The unit architecture in weeWX is designed to make basic unit conversions and display of units easy. It is not designed to provide dimensional analysis, arbitrary conversions, and indications of compatibility.

The driver reads observations from an instrument and converts them, as necessary, into a standard set of units. The actual units used by each instrument vary widely; some instruments use Metric units, others use US Customary units, and many use a mixture. The driver ensures that the units are consistent for storage in the weeWX database. By default, and to maintain compatibility with wview, the default database units are US Customary, although this can be changed.

Note that whatever unit system is used in the database, data can be displayed using any unit system. So, in practice, it does not matter what unit system is used in the database.

Each observation type, such as outTemp or pressure, is associated with a unit group such as group_temperature or group_pressure. Each unit group is associated with a unit type such as degree_F or mbar. The template system uses this architecture to display the names of units and to convert observations from one unit to another.

With this architecture one can easily create reports with, say, wind measured in knots, rain measured in mm, and temperatures in degree Celsius. Or one can create a single set of templates, but display data in different unit systems with only a few stanzas in a configuration file.

Customizing reports

This section discusses the two general strategies for customizing reports: by changing options in one or more configuration file, or by changing the template files. The former is generally easier, but occasionally the latter is necessary.

Changing options

Changing an option means either modifying the main configuration file weewx.conf, or the skin configuration file skin.conf.

Each skin will have a skin.conf that defines its default configuration. The examples in this guide refer to the standard skin that comes with the distribution.

Changing options in skin.conf

With this approach, edit the skin configuration file with a text editor. Changes made in this way will be used by weeWX the next time it generates reports, which is typically the next archive interval; there is no need to restart weeWX to see the results of the changes.

For the standard skin that comes with weeWX, the file is skins/Standard/skin.conf. It includes many, many options that can be changed. For a complete list, see the section below, The Standard skin.conf.

For example, suppose you wish to use metric units in the presentation layer, instead of the default US Customary Units. The section in skin.conf that controls units is [Units][[Groups]]. It looks like this:

[Units]
    [[Groups]]
        group_altitude    = foot
        group_degree_day  = degree_F_day
        group_direction   = degree_compass
        group_moisture    = centibar
        group_percent     = percent
        group_pressure    = inHg
        group_radiation   = watt_per_meter_squared
        group_rain        = inch
        group_rainrate    = inch_per_hour
        group_speed       = mile_per_hour
        group_speed2      = mile_per_hour2
        group_temperature = degree_F
        group_uv          = uv_index
        group_volt        = volt

To use metric units, you would edit this section to read:

[Units]
    [[Groups]]
        group_altitude    = meter
        group_degree_day  = degree_C_day
        group_direction   = degree_compass
        group_moisture    = centibar
        group_percent     = percent
        group_pressure    = mbar
        group_radiation   = watt_per_meter_squared
        group_rain        = mm
        group_rainrate    = mm_per_hour
        group_speed       = meter_per_second
        group_speed2      = meter_per_second2
        group_temperature = degree_C
        group_uv          = uv_index
        group_volt        = volt

The options that were changed have been  highlighted . Details of the various unit options are given in the Appendix: Units.

Other options are available, such as changing the text label for various observation types. For example, suppose your weather console is actually located in a barn, not indoors, and you want the plot for the temperature at the console to be labeled "Barn Temperature," rather than the default "Inside Temperature." This can be done by changing the inTemp option located in section [Labels][[Generic]] from the default

[Units]
    [[Generic]]
        inTemp  = Inside Temperature
        outTemp = Outside Temperature
        ...

to:

[Units]
    [[Generic]]
        inTemp  = Barn Temperature
        outTemp = Outside Temperature
        ...

Overriding options in skin.conf from weewx.conf

This approach is very similar, except that instead of changing the skin configuration file, skin.conf, directly, you override its options by editing the main configuration file, weewx.conf. The advantage of this approach is that you can use the same skin to produce several different output, each with separate options.

With this approach, you must restart weeWX to see the effects of any changes.

Revisiting our example, suppose you want two reports, one in US Customary, the other in Metric. The former will go in the directory HTML_ROOT, the latter in a directory, HTML_ROOT/metric. If you just simply modify skin.conf, you can get one, but not both at the same time. Alternatively, you could create a whole new skin by copying all the files to a new skin directory then editing the new skin.conf. The trouble with this approach is that you would then have two skins you would have to maintain. If you change something, you have to remember to change it in both places.

But, there's a better approach: reuse the same skin, but override some of its options. Here is what your [StdReport] section in weewx.conf would look like:

[StdReport]
    #
    # This section specifies what reports, using which skins, are to be generated.
    #

    # Where the skins reside, relative to WEEWX_ROOT:
    SKIN_ROOT = skins

    # Where the generated reports should go, relative to WEEWX_ROOT:
    HTML_ROOT = public_html

    # This report will use US Customary Units
    [[USReport]]
        # It is based on the Standard skin
        skin = Standard

    # This report will use metric units:
    [[MetricReport]]
        # It is also based on the Standard skin:
        skin = Standard
        # However, override where the results will go and put them in a directory:
        HTML_ROOT = public_html/metric

        # And override the options that were not in metric units
        [[[Units]]]
            [[[[Groups]]]]
                group_altitude    = meter
                group_pressure    = mbar
                group_rain        = mm
                group_rainrate    = mm_per_hour
                group_speed       = meter_per_second
                group_speed2      = meter_per_second2
                group_temperature = degree_C

We have done two things different from the stock reports. First (1), we have renamed the first report from StandardReport to USReport for clarity; and second (2), we have introduced a new report MetricReport, just like the first, except it puts its results in a different spot and uses different units. Both use the same skin, the Standard skin.

Customizing templates

If you cannot achieve the results you need by changing a configuration option, you may have to modify the templates that come with weeWX, or write your own.

Template modifications are preserved across upgrades (indeed, everything in the skins directory is preserved), so you don't have to worry about losing changes after an upgrade.

Template generation is done using the Cheetah templating engine. This is a very powerful engine, which essentially lets you have the full semantics of Python available in your templates. As this would make the templates incomprehensible to anyone but a Python programmer, weeWX adopts a very small subset of its power.

Tags

The key construct is a 'tag', specifying what value you want. For example:

$current.outTemp
$month.outTemp.max
$month.outTemp.maxtime

would code the current outside temperature, the maximum outside temperature for the month, and the time that maximum temperature occurred, respectively. So a template file that contains:

<html>
    <head>
        <title>Current conditions</title>
    </head>
    <body>
        <p>Current temperature = $current.outTemp</p>
        <p>Max for the month is $month.outTemp.max, which occurred at $month.outTemp.maxtime</p>
    </body>
</html>

would be all you need for a very simple HTML page that would display the text (assuming that the unit group for temperature is degree_F):

Current temperature = 51.0°F
Max for the month is 68.8°F, which occurred at 07-Oct-2009 15:15

The format that was used to format the temperature (51.0) is specified in section [Units][[StringFormat]]. The unit label °F is from section [Units][[Labels]], while the time format is from [Units][[TimeFormats]].

As we saw above, the tags can be very simple:

## Output max outside temperature using an appropriate format and label:
$month.outTemp.max

Most of the time, tags will "do the right thing" and are all you will need. However, weeWX offers extensive customization of the generated output for specialized applications such as XML RSS feeds, or ridgidly formatted reports (such as the NOAA reports). This section specifies the various options available.

There are two different versions of the tags, depending on whether the data is "current", or an aggregation over time. However, both versions are similar.

Time period $current

Time period $current represents a current observation. An example would be the current barometric pressure:

$current.barometer

Formally, weeWX first looks for the observation type in the record emitted by the NEW_ARCHIVE_RECORD event. This is generally the data emitted by the station console, augmented by any derived variables (e.g.wind chill) that you might have specified. If the observation type cannot be found there, the most recent record in the database will be searched.

The most general tag for a "current" observation looks like:

$current($data_binding=binding_name).obstype[.optional_unit_conversion][.optional_formatting]

Where:

binding_name is a binding name to a database. An example would be wx_binding. See the section Binding names for more details.

obstype is an observation type, such as barometer. See Appendix: Archive Types for a table of observation types valid for time period current.

optional_unit_conversion is an optional unit conversion tag. If provided, the results will be converted into the specified units, otherwise the default units specified in the skin configuration file (in section [Units][[Groups]]) will be used. See the section Unit Conversion Options.

optional_formatting is an optional formatting tag that controls how the value will appear. See the section Formatting Options.

Time period $latest

Time period $latest is very similar to $current, except that it uses the last available timestamp in a database. Usually, $current and $latest are the same, but if a data binding points to a remote database, they may not be. See the section Using multiple bindings for an example where this happened.

Aggregation periods

Aggregation periods is the other kind of tag. For example,

$week.rain.sum

represents an aggregation over time, using a certain aggregation type. In this example, the aggregation time is a week, and the aggregation type is summation. So, this tag represents the total rainfall over a week.

The most general tag for an aggregation over time looks like:

$period($data_binding=binding_name, $optional_ago=delta).statstype.aggregation[.optional_unit_conversion][.optional_formatting]

Where:

period is the time period over which the aggregation is to be done. Possible choices are listed in a table below.

binding_name is a binding name to a database. An example would be wx_binding. See the section Binding names for more details.

optional_ago is a keyword that depends on the aggregation period. For example, for week, it would be weeks_ago, for day, it would be days_ago, etc.

delta is an integer indicating which aggregation period is desired. For example $week($weeks_ago=1) indicates last week, $day($days_ago=2) would be the day-before-yesterday, etc. The default is zero: that is, this aggregation period.

statstype is a statistical type. This is generally any observation type that appears in the database, as well as a few synthetic types (such as heating and cooling degree-days). Not all aggregations are supported for all types.

aggregation is an aggregation type. If you ask for $month.outTemp.avg you are asking for the average outside temperature for the month. Possible aggregation types are given in Appendix: Aggregation types.

optional_unit_conversion is an optional unit conversion tag. If provided, the results will be converted into the specified units, otherwise the default units specified in the skin configuration file (in section [Units][[Groups]]) will be used. See the section Unit Conversion Options.

optional_formatting is an optional formatting tag that controls how the value will appear. See the section Formatting Options.

There are several different aggregation periods that can be used:

Aggregation period Meaning Example Meaning of example
$hour This hour. $hour.outTemp.maxtime The time of the max temperature this hour.
$day Today (since midnight). $day.outTemp.max The max temperature since midnight
$yesterday Yesterday. Synonym for $day($days_ago=1). $yesterday.outTemp.maxtime The time of the max temperature yesterday.
$week This week. The start of the week is set by option week_start. $week.outTemp.max The max temperature this week.
$month This month. $month.outTemp.min The minimum temperature this month.
$year This year. $year.outTemp.max The max temperature since the start of the year.
$rainyear This rain year. The start of the rain year is set by option rain_year_start. $rainyear.rain.sum The total rainfall for this rain year. The start of the rain year is set by option rain_year_start.

The $optional_ago parameters can be useful for statistics farther in the past. Here are some examples:

Aggregation period Example Meaning
$hour($hours_ago=h) $hour($hours_ago=1).outTemp.avg The average temperature last hour (1 hour ago).
$day($days_ago=d) $day($days_ago=2).outTemp.avg The average temperature day before yesterday (2 days ago).
$week($weeks_ago=d) $week($weeks_ago=1).outTemp.max The maximum temperature last week.
$month($months_ago=m) $month($months_ago=1).outTemp.max The maximum temperature last month.
$year($years_ago=m) $year($years_ago=1).outTemp.max The maximum temperature last year.

Unit conversion options

The tag optional_unit_conversion can be used with either current observations or aggregations. If supplied, the results will be converted to the specified units. For example, if you have set group_pressure to inches of mercury (inHg), then the tag

Today's average pressure=$day.barometer.avg 

would normally give a result such as

Today's average pressure=30.05 inHg

However, if you add mbar to the end,

$day.barometer.avg.mbar 

then the results will be in millibars:

Today's average pressure=1017.5 mbar

Wind ordinals

Using this method, you can output compass ordinals for wind direction. For example, the template

Current wind direction is $current.windDir ($current.windDir.ordinal_compass)

would result in:

Current wind direction is 138° (SW)

The ordinal abbreviations are set by option directions in the skin configuration file skin.conf.

Illegal conversions

If an inappropriate or nonsense conversion is asked for, e.g.,

Today's minimum pressure in mbars: $day.barometer.min.mbar
or in degrees C: $day.barometer.min.degree_C
or in foobar units: $day.barometer.min.foobar

then the offending tag(s) will be put in the output:

Today's minimum pressure in mbars: 1015.3
or in degrees C: $day.barometer.min.degree_C
or in foobar units: $day.barometer.min.foobar

Formatting options

The tag optional_formatting can be used with either current observations or aggregations. It can be one of:

Optional formatting tags
Optional formatting tag Comment
(no tag) Value is returned as a string, formatted using an appropriate string format from skin.conf. A unit label (e.g., °F) from skin.conf is also attached at the end.
.string(NONE_string) Value is returned as a string, formatted using an appropriate string format from skin.conf. If the value is None, the string NONE_string will be substituted if given, otherwise the value for NONE in [Units][[StringFormats]] will be used. A unit label (e.g., °F) from skin.conf will be attached at the end.
.formatted Value is returned as a string, formatted using an appropriate string format and None value from skin.conf. No unit label will be attached.
.format(string_formatNONE_string) Value is returned as a string, using the string format specified with string_format. If the value is None, the string NONE_string will be substituted if given, otherwise the value for NONE in [Units][[StringFormats]] will be used. A unit label (e.g., °F) from skin.conf will be attached at the end.
.nolabel(string_format, NONE_string) Value is returned as a string, using the string format specified with string_format. If the value is None, the string NONE_string will be substituted if given, otherwise the value for NONE in [Units][[StringFormats]] will be used. No unit label will be attached at the end.
.raw Value is returned "as is" without being converted to a string and without any formatting applied. This can be useful for doing arithmetic directly within the templates. You must be prepared to deal with a None value unless the value is converted directly to a string. In this case, it will be converted to the empty string ('')

 

Summary of formatting options
Formatting Tag Format Used Label Used NONE String Returned Value
(no tag) From skin.conf From skin.conf From skin.conf string
.string From skin.conf From skin.conf Optional user-supplied string
.formatted From skin.conf No label From skin.conf string
.format User-supplied From skin.conf Optional user-supplied string
.nolabel User-supplied No label Optional user-supplied string
.raw None No label None native value

Here are some examples with the expected results:

Formatting options with expected results
Tag Result Comment
$current.outTemp 45.2°F String formatting and label from skin.conf
$current.outTemp.string 45.2°F String formatting and label from skin.conf
$current.UV.string N/A This example assumes that the instrument has no UV sensor, resulting in a None value. The string specified by NONE in [Units][[StringFormats]] is substituted.
$current.UV.string("No UV") No UV This example assumes that the instrument has no UV sensor, resulting in a None value. The string supplied by the user is substituted.
$current.outTemp.formatted 45.2 String formatting from skin.conf; no label
$current.outTemp.format("%.3f") 45.200°F Specified string format used; label from skin.conf.
$current.dateTime 02-Apr-2010 16:25 Time formatting and label from skin.conf
$current.dateTime.format("%H:%M") 16:25 Specified time format used; label from skin.conf.
$current.dateTime.raw 1270250700 Unix epoch time, converted to string by template engine.
$current.outTemp.raw 45.2 Float returned, converted to string by template engine.
$month.dateTime 01-Apr-2010 00:00 Time formatting and label from skin.conf
$month.outTemp.avg 40.8°F String formatting and label from skin.conf
$month.outTemp.avg.string 40.8°F Time formatting and label from skin.conf
$month.UV.avg.string N/A This example assumes that the instrument has no UV sensor, resulting in a None value. The string specified by NONE in [Units][[StringFormats]] is substituted.
$month.UV.avg.string("No UV") No UV This example assumes that the instrument has no UV sensor, resulting in a None value. The string supplied by the user is substituted.
$month.outTemp.avg.formatted 40.8 String formatting from skin.conf; no label
$month.outTemp.avg.format("%.3f") 40.759°F Specified string format used; no label
$month.outTemp.avg.raw 40.7589690722 Float returned, converted to string by template engine
$month.UV.avg.raw (empty) None value converted to empty string by template engine.

Tags that take an argument, such as .string(NONE_string), do not require parenthesis if the argument is omitted. Thus, you can specify either $month.outTemp.string() or $month.outTemp.string, if you want the default value of NONE_string. They produce the same results.

Start, end, and dateTime

While not an observation type, in many ways the time of an observation, dateTime, can be treated as one. A tag such as

$current.dateTime

represents the current time (more properly, the time as of the end of the last archive interval) and would produce something like

01/09/2010 12:30:00

Like true observation types, explicit formats can be specified, except that they require a strftime() time format , rather than a string format.

For example, adding a format descriptor like this:

$current.dateTime.format("%d-%b-%Y %H:%M")

produces

09-Jan-2010 12:30

For aggregation periods, such as $month, you can request the start or end of the period, by using suffixes .start or .end, respectively. For example,

The current month runs from $month.start to $month.end.

results in

The current month runs from 01/01/2010 12:00:00 AM to 02/01/2017 12:00:00 AM.

In addition to the suffixes .start and .end, the suffix .dateTime is provided for backwards compatibility. Like .start, it refers to the start of the interval.

The returned string values will always be in local time. However, if you ask for the raw value

$current.dateTime.raw

the returned value will be in Unix Epoch Time (number of seconds since 00:00:00 UTC 1 Jan 1970, i.e., a large number), which you must convert yourself. It is guaranteed to never be None, so you don't worry have to worry about handling a None value.

Tag $trend

The tag $trend is available for time trends, such as changes in barometric pressure. Here are some examples:

Tag Results
$trend.barometer -.05 inHg
$trend($time_delta=3600).barometer -.02 inHg
$trend.outTemp 1.1 °C
$trend.time_delta 10800 secs
$trend.time_delta.hour 3 hrs

Note how you can explicitly specify a value in the tag itself (2nd example above). If you do not specify a value, then a default time interval, set by option time_delta in the skin configuration file, will be used. This value can be retrieved by using the syntax $trend.time_delta (3rd example above).

For example, the template expression

The barometer trend over $trend.time_delta.hour is $trend.barometer.format("%+.2f")

would result in

The barometer trend over 3 hrs is +.03 inHg.

Tag $span

The tag $span allows aggregation over a user defined period up to and including the current time. Its most general form looks like:

$span([$data_binding=binding_name][,$optional_delta=delta]).obstype.aggregation[.optional_unit_conversion][.optional_formatting]

Where:

binding_name is a binding name to a database. An example would be wx_binding. See the section Binding names for more details.

$optional_delta=delta is one or more comma separated delta settings from the table below. If more than one delta setting is included then the period used for the aggregate is the sum of the individual delta settings. If no delta setting is included, or all included delta settings are zero, the returned aggregate is based on the current obstype only.

obstype is a observation type, such as outTemp, that is supported by the $current tag. See Appendix: Archive Types for a table of observation types supported by the $current tag.

aggregation is an aggregation type. Possible aggregation types are given in Appendix: Aggregation types.

optional_unit_conversion is an optional unit conversion tag. See the section Unit Conversion Options.

optional_formatting is an optional formatting tag that controls how the value will appear. See the section Formatting Options.

There are several different delta settings that can be used:

Delta Setting Example Meaning
$time_delta=seconds $span($time_delta=1800).outTemp.avg The average temperature over the last immediate 30 minutes (1800 seconds).
$hour_delta=hours $span($hour_delta=6).outTemp.avg The average temperature over the last immediate 6 hours.
$day_delta=days $span($day_delta=1).rain.sum The total rainfall over the last immediate 24 hours.
$week_delta=weeks $span($week_delta=2).barometer.max The maximum barometric pressure over the last immediate 2 weeks.
$month_delta=months $span($month_delta=3).outTemp.min The minimum temperture over the last immediate 3 months (90 days).
$year_delta=years $span($year_delta=1).windchill.min The minimum wind chill over the last immediate 1 year (365 days).

For example, the template expressions

The total rainfall over the last 30 hours is $span($hour_delta=30).rain.sum

and

The total rainfall over the last 30 hours is $span($hour_delta=6, $day_delta=1).rain.sum

would both result in

The total rainfall over the last 30 hours is 1.24 in

Tag $unit

The type, label, and string formats for all units are also available, allowing you to do highly customized labels:

Tag Results
$unit.unit_type.outTemp degree_C
$unit.label.outTemp °C
$unit.format.outTemp %.1f

For example, the tag

$day.outTemp.max.formatted$unit.label.outTemp

would result in

21.2°C

(assuming metric values have been specified for group_temperature), essentially reproducing the results of the simpler tag $day.outTemp.max.

Tag $obs

The labels used for the various observation types are available using tag $obs. These are basically the values given in the skin dictionary, section [Labels][[Generic]].

Tag Results
$obs.label.outTemp Outside Temperature
$obs.label.UV UV Index

Iteration

It is possible to iterate over the following:

Tag suffix Results
.records Iterate over every record
.hours Iterate by hours
.days Iterate by days
.months Iterate by months
.years Iterate by years
.spans(interval=seconds) Iterate by custom length spans. The default interval is 10800 seconds (3 hours). The spans will align to local time boundaries.

The following template uses a Cheetah for loop to iterate over all months in a year, printing out each month's min and max temperature. The iteration loop is  highlighted .

Min, max temperatures by month
#for $month in $year.months
$month.dateTime.format("%B"): Min, max temperatures: $month.outTemp.min $month.outTemp.max
#end for
      

The result is:

Min, max temperatures by month:
January: Min, max temperatures: 30.1°F 51.5°F
February: Min, max temperatures: 24.4°F 58.6°F
March: Min, max temperatures: 27.3°F 64.1°F
April: Min, max temperatures: 33.2°F 52.5°F
May: Min, max temperatures: N/A N/A
June: Min, max temperatures: N/A N/A
July: Min, max temperatures: N/A N/A
August: Min, max temperatures: N/A N/A
September: Min, max temperatures: N/A N/A
October: Min, max temperatures: N/A N/A
November: Min, max temperatures: N/A N/A
December: Min, max temperatures: N/A N/A

The following template again uses a Cheetah for loop, this time to iterate over 3-hour spans over the last 24 hours, displaying the averages in each span. The iteration loop is  highlighted .

<p>3 hour averages over the last 24 hours</p>
<table>
  <tr>
    <td>Date/time</td><td>outTemp</td><td>outHumidity</td>
  </tr>
#for $_span in $span($day_delta=1).spans(interval=10800)
  <tr>
    <td>$_span.start.format("%d/%m %H:%M")</td><td>$_span.outTemp.avg</td><td>$_span.outHumidity.avg</td>
  </tr>
#end for
</table>      

The result is:

3 hour averages over the last 24 hours

Date/timeoutTempoutHumidity
21/01 18:5033.4°F95%
21/01 21:5032.8°F96%
22/01 00:5033.2°F96%
22/01 03:5033.2°F96%
22/01 06:5033.8°F96%
22/01 09:5036.8°F95%
22/01 12:5039.4°F91%
22/01 15:5035.4°F93%

See the NOAA template files NOAA/NOAA-YYYY.txt.tmpl and NOAA/NOAA-YYYY-MM.txt.tmpl for other examples using iteration, as well as explicit formatting.

Comprehensive example

This example is designed to put together a lot of the elements above, including iteration, aggregation period starts and ends, formatting, and overriding units.

<html>
  <head>
    <style>
      td { border: 1px solid #cccccc; padding: 5px; }
    </style>
  </head>

  <body>
    <table border=1 style="border-collapse:collapse;">
      <tr style="font-weight:bold">
    <td>Time interval</td>
    <td>Max temperature</td>
    <td>Time</td>
      </tr>
#for $hour in $day($days_ago=1).hours
      <tr>
    <td>$hour.start.format("%H:%M")-$hour.end.format("%H:%M")</td>
    <td>$hour.outTemp.max ($hour.outTemp.max.degree_C)</td>
    <td>$hour.outTemp.maxtime.format("%H:%M")</td>
      </tr>
#end for
      <caption>
    <p>
      Hourly max temperatures yesterday<br/>
      $day($days_ago=1).start.format("%d-%b-%Y")
    </p>
      </caption>
    </table>
  </body>
</html>

Click here for the results.

Almanac

If module pyephem has been installed, then weeWX can generate extensive almanac information for the Sun, Moon, Venus, Mars, Jupiter, and other heavenly bodies, including their rise, transit and set times, as well as their azimuth and altitude. Other information is also available.

Here is an example template:

Current time is $current.dateTime
#if $almanac.hasExtras
Sunrise, transit, sunset: $almanac.sun.rise $almanac.sun.transit $almanac.sun.set
Moonrise, transit, moonset: $almanac.moon.rise $almanac.moon.transit $almanac.moon.set
Mars rise, transit, set: $almanac.mars.rise $almanac.mars.transit $almanac.mars.set
Azimuth, altitude of mars: $almanac.mars.az $almanac.mars.alt
Next new, full moon: $almanac.next_new_moon $almanac.next_full_moon
Next summer, winter solstice: $almanac.next_summer_solstice $almanac.next_winter_solstice
#else
Sunrise, sunset: $almanac.sunrise $almanac.sunset
#end if

If pyephem is installed this would result in:

Current time is 29-Mar-2011 09:20
Sunrise, transit, sunset: 06:51 13:11 19:30
Moonrise, transit, moonset: 04:33 09:44 15:04
Mars rise, transit, set: 06:35 12:30 18:26
Azimuth, altitude of mars: 124.354959275 26.4808431952
Next new, full moon: 03-Apr-2011 07:32 17-Apr-2011 19:43
Next summer, winter solstice: 21-Jun-2011 10:16 21-Dec-2011 21:29

Otherwise, a fallback of basic calculations is used, resulting in:

Current time is 29-Mar-2011 09:20
Sunrise, sunset: 06:51 19:30

As shown in the example, you can test whether this extended almanac information is available with the value $almanac.hasExtras.

The almanac information falls in two categories:

We will cover each of these separately.

Calendar events

"Calendar events" do not require a heavenly body. They cover things such as next_solstice, or next_first_quarter_moon. The syntax here is

$almanac.next_solstice

or

$almanac.next_first_quarter_moon

Here is a table of the information that falls into this category:

Calendar events
previous_equinox next_equinox previous_solstice next_solstice
previous_autumnal_equinox next_autumnal_equinox previous_vernal_equinox next_vernal_equinox
previous_winter_solstice next_winter_solstice previous_summer_solstice next_summer_solstice
previous_new_moon next_new_moon previous_first_quarter_moon next_first_quarter_moon
previous_full_moon next_full_moon previous_last_quarter_moon next_last_quarter_moon

Heavenly bodies

The second category does require a heavenly body. This covers queries such as, "When does Jupiter rise?" or, "When does the sun transit?" Examples are

$almanac.jupiter.rise

or

$almanac.sun.transit

To accurately calculate these times, weeWX automatically uses the present temperature and pressure to calculate refraction effects. However, you can override these values, which will be necessary if you wish to match the almanac times published by the Naval Observatory as explained in the pyephem documentation. For example, to match the sunrise time as published by the Observatory, instead of

$almanac.sun.rise

use

$almanac(pressure=0, horizon=-34.0/60.0).sun.rise

By setting pressure to zero we are bypassing the refraction calculations and manually setting the horizon to be 34 arcminutes lower than the normal horizon. This is what the Navy uses.

If you wish to calculate the start of civil twilight, you can set the horizon to -6 degrees, and also tell weeWX to use the center of the sun (instead of the upper limb, which it normally uses) to do the calcuation:

$almanac(pressure=0, horizon=-6).sun(use_center=1).rise

The general syntax is:

$almanac(pressure=pressure, horizon=horizon,
         temperature=temperature_C).heavenly_body(use_center=[01]).attribute
      

As you can see, in addition to the horizon angle, you can also override atmospheric pressure and temperature (degrees Celsius).

PyEphem offers an extensive list of objects that can be used for the heavenly_body tag. All the planets and many stars are in the list.

The possible values for the attribute tag are listed in the following table:

Attributes that can be used with heavenly bodies
az alt a_ra a_dec
g_ra ra g_dec dec
elong radius hlong hlat
sublat sublong next_rising next_setting
next_transit next_antitransit previous_rising previous_setting
previous_transit previous_antitransit rise set
transit      

Wind

Wind deserves a few comments because it is stored in the database in two different ways: as a set of scalars, and as a vector of speed and direction. Here are the four wind-related scalars stored in the main archive database:

Archive type Meaning Valid contexts
windSpeed The average wind speed seen during the archive period. $current, $latest, $day, $week, $month, $year, $rainyear
windDir If software record generation is used, this is the vector average over the archive period. If hardware record generation is used, the value is hardware dependent.
windGust The maximum (gust) wind speed seen during the archive period.
windGustDir The direction of the wind when the gust was observed.

In addition, a wind vector is stored in the daily summaries.

Daily summary type Meaning Valid contexts
wind A vector composite of the wind. It includes information such as the direction of the maximum gust, and the x- and y-vector wind run. $day, $week, $month, $year, $rainyear

Note
The vector is only stored in the daily summaries, so unlike the scalar tags such as windSpeed or windGust, the tag wind can only be used in aggregations such as $day, $month, etc.

Any of these can be used in your tags. Here are some examples:

Tag Meaning
$current.windSpeed The average wind speed over the most recent archive interval.
$current.windDir If software record generation is used, this is the vector average over the archive interval. If hardware record generation is used, the value is hardware dependent.
$current.windGust The maximum wind speed (gust) over the most recent archive interval.
$current.windGustDir The direction of the gust.
$day.windSpeed.avg The average wind speed since midnight. If the wind blows east at 5 m/s for 2 hours, then west at 5 m/s for 2 hours, the average wind speed is 5 m/s.
$day.wind.avg The average wind speed since midnight. Same as $day.windSpeed.avg above.
$day.wind.vecavg The vector average wind speed since midnight. If the wind blows east at 5 m/s for 2 hours, then west at 5 m/s for 2 hours, the vector average wind speed is zero.
$day.windSpeed.max The max average wind speed. The wind is averaged over each of the archive intervals. Then the maximum of these values is taken. Note that this is not the same as the maximum observed wind speed.
$day.windGust.max The maximum observed wind speed since midnight, i.e., the maximum gust.
$day.windDir.avg Not a very useful quantity. This is the strict, arithmetic average of all the compass wind direction. Probably not what you want.
$day.wind.vecdir The direction of the vector averaged wind speed. If the wind blows northwest for two hours, then southwest for two hours, the vector averaged direction is west.

Defining new tags

We have seen how you can change a template and make use of the various tags available such as $day.outTemp.max for the maximum outside temperature for the day. But, what if you want to introduce some new data for which no tag is available?

If you wish to introduce a static tag, that is, one that will not change with time (such as a Google analytics Tracker ID, or your name), then this is very easy: simply put it in section [Extras] in the skin configuration file. More information on how to do this can be found there.

But, what if you wish to introduce a more dynamic tag, one that requires some calculation, or perhaps uses the database? Simply putting it in the [Extras] section won't do, because then it cannot change.

The answer is to write a search list extension.

How the search list works

Let's start by taking a look at how the Cheetah search list works.

The Cheetah template engine finds tags by scanning a search list, a Python list of objects. For example, for a tag $foo, the engine will scan down the list, trying each object in the list in turn. For each object, it will first try using foo as an attribute, that is, it will try evaluating obj.foo. If that raises an AttributeError exception, then it will try foo as a key, that is obj[key]. If that raises a KeyError exception, then it moves on to the next item in the list. The first match that does not raise an exception is used. If no match is found, Cheetah raises a NameMapper.NotFound exception.

How tags work

Now let's take a look at how the search list interacts with weeWX tags. Let's start by looking at a simple example: station altitude, available as the tag

$station.altitude

As we saw in the previous section, Cheetah will run down the search list, looking for an object with a key or attribute station. In the default search list, weeWX includes one such object, an instance of the class weewx.cheetahgenerator.Station, which has an attribute station, so it gets a hit on this object.

Cheetah will then try to evaluate the attribute altitude on this object. Class Station has such an attribute, so Cheetah evaluates it.

What this attribute returns is not a raw value, say 700, nor even a string. Instead, it returns an instance of the class ValueHelper, a special class defined in module weewx.units. Internally, it holds not only the raw value, but also references to the formats, labels, and conversion targets you specified in your configuration file. Its job is to make sure that the final output reflects these preferences. Cheetah doesn't know anything about this class. What it needs, when it has finished evaluating the expression $station.altitude, is a string. In order to convert the ValueHelper it has in hand into a string, it does what every other Python object does when faced with this problem: it calls the special method __str__. Class ValueHelper has a definition for this method. Evaluating this function triggers the final steps in this process. Any necessary unit conversions are done, then formatting occurs and, finally, a label is attached. The result is a string something like

700 feet

which is what Cheetah actually puts in the generated HTML file. This is a good example of lazy evaluation. The tags gather all the information they need, but don't do the final evaluation until the last final moment, when the most context is understood. WeeWX uses this technique extensively.

Now let's look at a more complicated example, say the maximum temperature since midnight:

$day.outTemp.max

When this is evaluated by Cheetah, it actually produces a chain of objects. At the top of this chain is class weewx.tags.TimeBinder, an instance of which is included in the default search list. Internally, this instance stores the time of the desired report (usually the time of the last archive record), a cache to the databases, a default data binding, as well as references to the formatting and labelling options you have chosen.

This instance is examined by Cheetah to see if it has an attribute day. It does and, when it is evaluated, it returns the next class in the chain, an instance of weewx.tags.TimespanBinder. In addition to all the other things contained in its parent TimeBinder, class TimespanBinder adds the desired time period, that is, the time span from midnight to the current time.

Cheetah then continues on down the chain and tries to find the next attribute, outTemp. There is no such hard coded attribute (hard coding all the conceivable different observation types would be impossible!). Instead, class TimespanBinder defines the Python special method __getattr__. If Python cannot find a hard coded version of an attribute, and the method __getattr__ exists, it will try it. The definition provided by TimespanBinder returns an instance of the next class in the chain, weewx.tags.ObservationBinder, which not only remembers all the previous stuff, but also adds the observation type, outTemp.

Cheetah then tries to evaluate an attribute max of this class. Now, finally, the chain ends. The attribute max triggers the actual calculation of the value, using all the known parameters: the database binding to be hit, the time span of interest, the observation type, and the type of aggregation, querying the database as necessary. The database is not actually hit until the last possible moment, after everything needed to do the evalation is known.

Like our previous example, the results of the evaluation are then packaged up in an instance of ValueHelper, which does the final conversion to the desired units, formats the string, then adds a label. The results, something like

12°C

are put in the generated HTML file. As you can see, a lot of machinery is hidden behind the deceptively simple expression $day.outTemp.max!

Extending the list

As mentioned, weeWX comes with a number of objects already in the search list, but you can extend it. To do so, you should have some familiarity with Python, in particular, how to write new classes and member functions for them.

Let's look at an example. The regular version of weeWX offers statistical summaries by day, week, month, and year. Suppose we would like to add two more:

This example is included in the distribution as examples/stats.py:

import datetime
import time

from weewx.cheetahgenerator import SearchList
from weewx.tags import TimespanBinder
from weeutil.weeutil import TimeSpan

class MyStats(SearchList):                                                   # 1
    """My search list extension"""

    def __init__(self, generator):                                           # 2
        SearchList.__init__(self, generator)

    def get_extension_list(self, timespan, db_lookup):                       # 3
        """Returns a search list extension with two additions.

        Parameters:
          timespan: An instance of weeutil.weeutil.TimeSpan. This will
                    hold the start and stop times of the domain of
                    valid times.

          db_lookup: This is a function that, given a data binding
                     as its only parameter, will return a database manager
                     object.
        """

        # First, create TimespanBinder object for all time. This one is easy
        # because the object timespan already holds all valid times to be
        # used in the report.
        all_stats = TimespanBinder(timespan,
                                   db_lookup,
                                   formatter=self.generator.formatter,
                                   converter=self.generator.converter)       # 4

        # Now get a TimespanBinder object for the last seven days. This one we
        # will have to calculate. First, calculate the time at midnight, seven
        # days ago. The variable week_dt will be an instance of datetime.date.
        week_dt = datetime.date.fromtimestamp(timespan.stop) - \
                    datetime.timedelta(weeks=1)                              # 5
        # Convert it to unix epoch time:
        week_ts = time.mktime(week_dt.timetuple())                           # 6
        # Form a TimespanBinder object, using the time span we just
        # calculated:
        seven_day_stats = TimespanBinder(TimeSpan(week_ts, timespan.stop),
                                         db_lookup,
                                         formatter=self.generator.formatter,
                                         converter=self.generator.converter) # 7

        # Now create a small dictionary with keys 'alltime' and 'seven_day':
        search_list_extension = {'alltime'   : all_stats,
                                 'seven_day' : seven_day_stats}              # 8

        # Finally, return our extension as a list:
        return [search_list_extension]                                       # 9

Going through the example, line by line:

  1. Create a new class called MyStats, which will inherit from class SearchList. All search list extensions inherit from this class.
  2. Create an initializer for our new class. In this case, the initializer is not really necessary and does nothing except pass its only parameter, generator, a reference to the calling generator, on to its superclass, SearchList. The superclass will store it in self.
  3. Override member function get_extension_list(). This function will be called when the generator is ready to accept your new search list extension. The parameters that will be passed in are:
    • self Python's way of indicating the instance we are working with;
    • timespan An instance of the utility class TimeSpan. This will contain the valid start and ending times used by the template. Normally, this is all valid times;
    • db_lookup This is a function supplied by the generator. It takes a single argument, a name of a binding. When called, it will return an instance of the database manager class for that binding. The default for the function is whatever binding you set with the option data_binding for this report, usually wx_binding.
  4. The class TimespanBinding represents a statistical calculation over a time period. We have already met it in the introduction How tags work. In our case, we will set it up to represent the statistics over all possible times. The class takes 4 parameters.
    • The first is the timespan over which the calculation is to be done. Here, we have a lucky coincidence: the variable timespan already holds a TimeSpan object representing the domain of all valid timespans, so we simply pass it in.
    • The second is the database lookup function to be used. We simply pass in db_lookup.
    • The third should be an instance of class weewx.units.Formatter, which contains information about how the results should be formatted. We just pass in the formatter set up by the generator, self.generator.formatter.
    • The fourth should be an instance of weewx.units.Converter, which contains information about the target units (e.g., degree_C) that are to be used. Again, we just pass in the instance set up by the generator, self.generator.converter.

That one was relatively easy because we already had an instance of TimeSpan, that is, timespan, which represented the time over which we wanted to do the calculations. Setting up an instance that will work for the last seven days is a bit trickier. Continuing our example...

  1. The object timespan holds the domain of all valid times, but in order to calculate statistics for the last seven days, we need not the earliest valid time, but the time at midnight seven days ago. So, we do a little Python date arithmetic to calculate this. The object week_dt will be an instance of datetime.date.
  2. We convert it to unix epoch time.
  3. Now we are ready to initialize an appropriate TimespanBinder object. It's the same as in step #4, except we use our new timespan object.
  4. Create a small dictionary with two keys, alltime, and seven_day.
  5. Return the dictionary in a list

The final step that we need to do is to tell the template engine where to find our extension. You do that by going into the skin configuration file, skin.conf, and adding the option search_list_extensions with our new extension. When you're done, it will look something like this:

[CheetahGenerator]
    # This section is used by the generator CheetahGenerator, and specifies
    # which files are to be generated from which template.

    # Possible encodings are 'html_entities', 'utf8', or 'strict_ascii'
    encoding = html_entities
    search_list_extensions = user.stats.MyStats

    [[SummaryByMonth]]
    ...

Our addition has been highlighted. Note that it is in the section [CheetahGenerator].

Now, if the Cheetah engine encounters the tag $alltime, it will scan the search list, looking for an attribute or key that matches alltime. When it gets to the little dictionary we provided, it will find a matching key, allowing it to retrieve the appropriate TimespanBinding object.

With this approach, you can now include "all time" or "seven day" statistics in your HTML templates:

<table>
    <tr>
        <td>Maximum temperature to date: </td>
        <td>$alltime.outTemp.max</td>
    </tr>
    <tr>
        <td>Minimum temperature to date: </td>
        <td>$alltime.outTemp.min
    </tr>
    <tr>
        <td>Rain over the last seven days: </td>
        <td>$seven_day.rain.sum
    </tr>
</table>

If you place a custom generator somewhere other than the hierarchy where weewxd resides, you may have to specify its location in the environment variable PYTHONPATH in the shell where you start weeWX:

export PYTHONPATH=/home/me/secret_location

Customizing images

The installed version of weeWX is configured to generate a set of useful plots. But, what if you don't like how they look, or you want to generate different plots, perhaps with different aggregation types? This section covers how to do this.

Image generation is controlled by the section [ImageGenerator] in the skin configuration file skin.conf. Let's take a look at the beginning part of this section. It looks like this:

[ImageGenerator]
    ...
    image_width = 300
    image_height = 180
    image_background_color = 0xf5f5f5

    chart_background_color = 0xd8d8d8
    chart_gridline_color = 0xa0a0a0
    ...

The options right under the section name [ImageGenerator] will apply to all plots, unless overridden in subsections. So, unless otherwise changed, all plots will be 300 pixels in width, 180 pixels in height, and will have an RGB background color of 0xf5f5f5, a very light gray (HTML color "WhiteSmoke"). The chart itself will have a background color of 0xd8d8d8 (a little darker gray), and the gridlines will be 0xa0a0a0 (still darker). The other options farther down (not shown) will also apply to all plots.

Time periods

After the "global" options at the top of section [ImageGenerator], comes a set of sub-sections, one for each time period (day, week, month, and year). These sub-sections define the nature of aggregation and plot types for that time period. For example, here is a typical set of options for sub-section [[month_images]]. It controls which "monthly" images will get generated, and what they will look like:

    [[month_images]]
        x_label_format = %d
        bottom_label_format = %m/%d/%y %H:%M
        time_length = 2592000    # == 30 days
        aggregate_type = avg
        aggregate_interval = 10800    # == 3 hours
        show_daynight = false

The option x_label_format gives a strftime() type format for the x-axis. In this example, it will only show days (format option %d). The bottom_label_format is the format used to time stamp the image at the bottom. In this example, it will show the time as something like 10/25/09 15:35. A plot will cover a nominal 30 days, and all items included in it will use an aggregate type of averaging over 3 hours. Finally, by setting option show_daynight to false, we are requesting that day-night, shaded bands not be shown.

Image files

Within each time period sub-section is another nesting, one for each image to be generated. The title of each sub-sub-section is the filename to be used for the image. Finally, at one additional nesting level (!) are the logical names of all the line types to be drawn in the image. Like elsewhere, the values specified in the level above can be overridden. For example, here is a typical set of options for sub-sub-section [[[monthrain]]]:

        [[[monthrain]]]
            plot_type = bar
            yscale = None, None, 0.02
            [[[[rain]]]]
                aggregate_type = sum
                aggregate_interval = 86400
                label = Rain (daily total)

This will generate an image file with name monthrain.png. It will be a bar plot. Option yscale controls the y-axis scaling — if left out, the scale will automatically be chosen. However, in this example we are choosing to exercise some degree of control by specifying values explicitly. The option is a 3-way tuple (ylow, yhigh, min_interval), where ylow and yhigh are the minimum and maximum y-axis values, respectively, and min_interval is the minimum tick interval. If set to None, the corresponding value will be automatically chosen. So, in this example, the setting

yscale = None, None, 0.02

will cause weeWX to pick sensible y minimum and maximum values, but require that the tick increment (min_interval) be at least 0.02.

Continuing on with the example above, there will be only one plot "line" (it will actually be a series of bars) and it will have logical name rain. Because we have not said otherwise, the SQL data type to be used for this line will be the same as its logical name, that is, rain, but this can be overridden. The aggregation type will be summing (overriding the averaging specified in sub-section [[month_images]]), so you get the total rain over the aggregate period (rather than the average) over an aggregation interval of 86,400 seconds (one day). The plot line will be titled with the indicated label of 'Rain (daily total)'. The result of all this is the following plot:

Sample monthly rain plot

Line gaps

If there is a time gap in the data, the option line_gap_fraction controls how line plots will be drawn. Here's what a plot looks like without and with this option being specified:

Gap not shown
No line_gap_fraction specified
Gap showing
With line_gap_fraction=0.01

Including more than one SQL type in a plot

More than one SQL type can be included in a plot. For example, here is how to generate a plot with the week's outside temperature as well as dewpoint:

[[[monthtempdew]]]
    [[[[outTemp]]]]
    [[[[dewpoint]]]]

This would create an image in file monthtempdew.png that includes a line plot of both outside temperature and dewpoint.

Including the same SQL type more than once in a plot

Another example. Say you want a plot of the day's temperature, overlaid with hourly averages. Here, you are using the same data type (outTemp) for both plot lines, the first with averages, the second without. If you do the obvious it won't work:

## WRONG ##
[[[daytemp_with_avg]]]
    [[[[outTemp]]]]
        aggregate_type = avg
        aggregate_interval = 3600
    [[[[outTemp]]]]  # OOPS! The same section name appears more than once!

The option parser does not allow the same section name (outTemp in this case) to appear more than once at a given level in the configuration file, so an error will be declared (technical reason: formally, the sections are an unordered dictionary). If you wish for the same SQL type to appear more than once in a plot then there is a trick you must know: use option data_type. This will override the default action that the logical line name is used for the SQL type. So, our example would look like this:

[[[daytemp_with_avg]]]
    [[[[avgTemp]]]]
        data_type = outTemp
        aggregate_type = avg
        aggregate_interval = 3600
        label = Avg. Temp.
    [[[[outTemp]]]]

Here, the first plot line has been given the name avgTemp to distinguish it from the second line outTemp. Any name will do — it just has to be different. We have specified that the first line will use data type outTemp and that it will use averaging over a one hour period. The second also uses outTemp, but will not use averaging.

The result is a nice plot of the day's temperature, overlaid with a one hour smoothed average:

Daytime temperature with running average

One more example. This one shows daily high and low temperatures for a year:

[[year_images]]
    [[[yearhilow]]]
        [[[[hi]]]]
            data_type = outTemp
            aggregate_type = max
            label = High
        [[[[low]]]]
            data_type = outTemp
            aggregate_type = min
            label = Low Temperature

This results in the plot yearhilow.png:

Daily highs and lows

Progressive vector plots

WeeWX can produce progressive vector plots as well as the more conventional x-y plots. To produce these, use plot type vector. You need a vector type to produce this kind of plot. There are two: windvec, and windgustvec. While they do not actually appear in the SQL database, weeWX understands that they represent special vector-types. The first, windvec, represents the average wind in an archive period, the second, windgustvec the max wind in an archive period. Here's how to produce a progressive vector for one week that shows the hourly biggest wind gusts, along with hourly averages:

[[[weekgustoverlay]]]
    aggregate_interval = 3600
    [[[[windvec]]]]
        label = Hourly Wind
        plot_type = vector
        aggregate_type = avg
    [[[[windgustvec]]]]
        label = Gust Wind
        plot_type = vector
        aggregate_type = max

This will produce an image file with name weekgustoverlay.png. It will consist of two progressive vector plots, both using hourly aggregation (3,600 seconds). For the first set of vectors, the hourly average will be used. In the second, the max of the gusts will be used:

hourly average wind vector overlaid with gust vectors

By default, the sticks in the progressive wind plots point towards the wind source. That is, the stick for a wind from the west will point left. If you have a chronic wind direction (as I do), you may want to rotate the default direction so that all the vectors do not line up over the x-axis, overlaying each other. Do this by using option vector_rotate. For example, with my chronic westerlies, I set vector_rotate to 90.0 for the plot above, so winds out of the west point straight up.

If you use this kind of plot (the out-of-the-box version of weeWX includes daily, weekly, monthly, and yearly progressive wind plots), a small compass rose will be put in the lower-left corner of the image to show the orientation of North.

Overriding values

Remember that values at any level can override values specified at a higher level. For example, say you want to generate the standard plots, but for a few key observation types such as barometer, you want to also generate some oversized plots to give you extra detail, perhaps for an HTML popup. The standard weewx.conf file specifies plot size of 300x180 pixels, which will be used for all plots unless overridden:

[ImageGenerator]
    ...
    image_width = 300
    image_height = 180

The standard plot of barometric pressure will appear in daybarometer.png:

[[[daybarometer]]]
    [[[[barometer]]]] 

We now add our special plot of barometric pressure, but specify a larger image size. This image will be put in file daybarometer_big.png.

[[[daybarometer_big]]]
    image_width  = 600
    image_height = 360
    [[[[barometer]]]]

Using multiple bindings

It's easy to use more than one database in your reports. Here's an example. In my office I have two consoles: a VantagePro2 connected to a Dell Optiplex, and a WMR100N, connected to a Raspberry Pi. Each is running weeWX. The Dell is using SQLite, the RPi, MySQL.

Suppose I wish to compare the inside temperatures of the two consoles. How would I do that?

It's easier to access MySQL across a network than SQLite, so let's run the reports on the Dell, but access the RPi's MySQL database remotely. Here's how the bindings and database sections of weewx.conf would look on the Dell:

[DataBindings]
    # This section binds a data store to an actual database

    [[wx_binding]]
        # The database to be used - it should match one of the sections in [Databases]
        database = archive_sqlite
        # The name of the table within the database
        table_name = archive
        # The class to manage the database
        manager = weewx.wxmanager.WXDaySummaryManager
        # The schema defines to structure of the database contents
        schema = schemas.wview.schema

    [[wmr100_binding]]
        # Binding for my WMR100 on the RPi
        database = rpi_mysql
        # The name of the table within the database
        table_name = archive
        # The class to manage the database
        manager = weewx.wxmanager.WXDaySummaryManager
        # The schema defines to structure of the database contents
        schema = schemas.wview.schema

[Databases]
    # This section binds to the actual database to be used

    [[archive_sqlite]]
        database_type = SQLite
        database_name = weewx.sdb

    [[rpi_mysql]]
        database_type = MySQL
        database_name = weewx
        host = rpi-bug

[DatabaseTypes]
    #   This section defines defaults for the different types of databases.

    [[SQLite]]
        driver = weedb.sqlite
        # Directory in which the database files are located
        SQLITE_ROOT = %(WEEWX_ROOT)s/archive

    [[MySQL]]
        driver = weedb.mysql
        # The host where the database is located
        host = localhost
        # The user name for logging in to the host
        user = weewx
        # The password for the user name
        password = weewx
    

The two additions have been highlighted. The first, [[wmr100_binding]], adds a new binding called wmr10_binding. It links ("binds") to the new database, called rpi_mysql, through the option database. It also defines some characteristics of the binding, such as which manager is to be used and what its schema looks like.

The second addition, [[rpi-mysql]] defines the new database. Option database_type is set to MySQL, indicating that it is a MySQL database. Defaults for MySQL databases are defined in the section [[MySQL]]. The new database accepts all of them, except for host, which as been set to the remote host rpi-bug, the name of my Raspberry Pi.

Explicit binding in tags

How do we use this new binding? First, let's do a text comparison, using tags. Here's what our template looks like:

<table>
  <tr>
    <td class="stats_label">Inside Temperature, Vantage</td>
    <td class="stats_data">$current.inTemp</td>
  </tr>
  <tr>
    <td class="stats_label">Inside Temperature, WMR100</td>
    <td class="stats_data">$latest($data_binding='wmr100_binding').inTemp</td>
  </tr>
</table>

The explicit binding to wmr100_binding is highlighted. This tells the reporting engine to override the default binding specifed in [StdReport], generally wx_binding, and use wmr100_binding instead.

This results in an HTML output that looks like:

Inside Temperature, Vantage 68.7°F
Inside Temperature, WMR100 68.9°F

Explicit binding in images

How would we produce a graph of the two different temperatures? Here's what the relevant section of the skin.conf file would look like.

[[[daycompare]]]
   [[[[inTemp]]]]
       label = Vantage inTemp
       [[[[WMR100Temp]]]]
           data_type = inTemp
           data_binding = wmr100_binding
           label = WMR100 inTemp

This will produce an image with name daycompare.png, with two plot lines. The first will be of the temperature from the Vantage. It uses the default binding, wx_binding, and will be labeled Vantage inTemp. The second explicitly uses the wmr100_binding. Because it uses the same variable name (inTemp) as the first line, we had to explicitly specify it using option data_type, in order to avoid using the same sub-section name twice (see the section Including the same SQL type more than once in a plot for details). It will be labeled WMR100 inTemp. The results look like this:

Comparing temperatures

Stupid detail

At first, I could not get this example to work. The problem turned out to be that the RPi was processing things just a beat behind the Dell, so the temperature for the "current" time wasn't ready when the Dell needed it. I kept getting N/A. To avoid this, I introduced the tag $latest, which uses the last available timestamp in the binding, which may or may not be the same as what $current uses. That's why the example above uses $latest instead of $current.

Customizing the report generation time

Normal weeWX operation is to run each report defined in weewx.conf every archive period. Whilst this may suit most situations, there may be occasions when it is desirable to run a report less frequently than every archive period. For example, the archive interval might be 5 minutes, but you only want to FTP files every 30 minutes, or once per day. In such cases the report_timing option can be used to control when individual reports are run.

Note
Whilst the report_timing option allows the user to specify when a given report is generated, it should be noted that the generation of reports is still controlled by the weeWX report cycle and as such the report_timing option does not allow reports to be generated more frequently than once every archive period.

The report_timing option

The report_timing option uses a CRON-like format to control when a report is to be run. Whilst a CRON-like format is used, the control of weeWX report generation using the report_timing option is confined completely to weeWX and has no interraction with the system CRON service.

The report_timing option consists of five, space separated parameters as follows:

report_timing = minutes hours day_of_month months day_of_week

The report_timing parameters are summarised in the following table:

Parameter Function Allowable values
minutes Specifies the minutes of the hour when the report will be run *, or
numbers in the range 0..59 inclusive
hours Specifies the hours of the day when the report will be run *, or
numbers in the range 0..23 inclusive
day_of_month Specifies the days of the month when the report will be run *, or
numbers in the range 1..31 inclusive
months Specifies the months of the year when the report will be run *, or
numbers in the range 1..12 inclusive, or
abbreviated names in the range jan..dec inclusive
day_of_week Specifies the days of the week when the report will be run *, or
numbers in the range 0..7 inclusive (0,7 = Sunday, 1 = Monday etc), or
abbreviated names in the range sun..sat inclusive

The report_timing option may only be used in weewx.conf. When set in the [StdReport] section of weewx.conf the option will apply to all reports listed under [StdReport]. When specified within a report section, the option will override any setting in [StdReport] for that report. In this manner it is possible to have different reports run at different times. The following sample weewx.conf excerpt illustrates this:

[StdReport]

    # Where the skins reside, relative to WEEWX_ROOT
    SKIN_ROOT = skins

    # Where the generated reports should go, relative to WEEWX_ROOT
    HTML_ROOT = public_html

    # The database binding indicates which data should be used in reports.
    data_binding = wx_binding

    # Report timing parameter
    report_timing = 0 * * * *

    # Each of the following subsections defines a report that will be run.

    [[AReport]]
        skin = SomeSkin

    [[AnotherReport]]
        skin = SomeOtherSkin
        report_timing = */10 * * * *

In the above case, the [[AReport]] report would be run under under control of the 0 * * * * setting (on the hour) under [StdReport] and the [[AnotherReport]] report would be run under control of the */10 * * * * setting (every 10 minutes) which has overriden the [StdReport] setting.

How report_timing controls reporting

The syntax and interpretation of the report_timing parameters are largely the same as those of the CRON service in many Unix and Unix-like operating systems. The syntax and interpretation are outlined below.

When the report_timing option is in use weeWX will run a report when the minute, hour and month of year parameters match the report time, and at least one of the two day parameters (day of month or day of week) match the report time. This means that non-existent times, such as "missing hours" during daylight savings changeover, will never match, causing reports scheduled during the "missing times" not to be run. Similarly, times that occur more than once (again, during daylight savings changeover) will cause matching reports to be run more than once.

Note
Report time does not refer to the time at which the report is run, but rather the date and time of the latest data the report is based upon. If you like, it is the effective date and time of the report. For normal weeWX operation, the report time aligns with the dateTime of the most recent archive record. When reports are run using the wee_reports utility, the report time is either the dateTime of the most recent archive record (the default) or the optional timestamp command line argument.

Note
The day a report is to be run can be specified by two parameters; day of month and/or day of week. If both parameters are restricted (i.e., not an asterisk), the report will be run when either field matches the current time. For example,
report_timing = 30 4 1,15 * 5
would cause the report to be run at 4:30am on the 1st and 15th of each month as well as 4:30am every Friday.

The relationship between report_timing and archive period

A traditional CRON service has a resolution of one minute, meaning that the CRON service checks each minute as to whether to execute any commands. On the other hand, the weeWX report system checks which reports are to be run once per archive period, where the archive period may be one minute, five minutes or some other user defined period. Consequently, the report_timing option may specify a report to be run at some time that does not align with the weeWX archive period. In such cases cases the report_timing option does not cause a report to be run outside of the normal weeWX report cycle, rather it will cause the report to be run during the next report cycle. At the start of each report cycle, and provided a report_timing option is set, weeWX will check each minute boundary from the current report time back until the report time of the previous report cycle. If a match is found on any of these one minute boundaries the report will be run during the report cycle. This may be best described through some examples:

report_timing Archive period When the report will be run
0 * * * * 5 minutes The report will be run only during the report cycle commencing on the hour.
5 * * * * 5 minutes The report will be run only during the report cycle commencing at 5 minutes past the hour.
3 * * * * 5 minutes The report will be run only during the report cycle commencing at 5 minutes past the hour.
10 * * * * 15 minutes The report will be run only during the report cycle commencing at 15 minutes past the hour
10,40 * * * * 15 minutes The report will be run only during the report cycles commencing at 15 minutes past the hour and 45 minutes past the hour.
5,10 * * * * 15 minutes The report will be run once only during the report cycle commencing at 15 minutes past the hour.

Lists, ranges and steps

The report_timing option supports lists, ranges and steps for all parameters. Lists, ranges and steps may be used as follows:

Nicknames

The report_timing option supports a number of time specification 'nicknames'. These nicknames are prefixed by the '@' character and replace the five parameters in the report_timing option. The nicknames supported are:

Nickname Equivalent setting When the report will be run
@yearly
@annually
0 0 1 1 * Once per year at midnight on 1 January.
@monthly 0 0 1 * * Monthly at midnight on the 1st of the month.
@weekly 0 0 * * 0 Every week at midnight on Sunday.
@daily 0 0 * * * Every day at midnight.
@hourly 0 * * * * Every hour on the hour.

report_timing option examples

Numeric settings for report_timing can be at times difficult to understand due to the complex combinations of parameters. The following table shows a number of example report_timing options and the corresponding times when the report would be run.

report_timing When the report will be run
* * * * * Every archive period. This setting is effectively the default weeWX method of operation.
25 * * * * 25 minutes past every hour.
0 * * * * Every hour on the hour.
5 0 * * * 00:05 daily.
25 16 * * * 16:25 daily.
25 16 1 * * 16:25 on the 1st of each month.
25 16 1 2 * 16:25 on the 1st of February.
25 16 * * 0 16:25 each Sunday.
*/10 * * * * On the hour and 10, 20, 30, 40 and 50 mnutes past the hour.
*/9 * * * * On the hour and 9, 18, 27, 36, 45 and 54 minutes past the hour.
*/10 */2 * * * 0, 10, 20, 30, 40 and 50 minutes after the even hour.
* 6-17 * * * Every archive period from 06:00 (inclusive) up until, but excluding, 18:00.
* 1,4,14 * * * Every archive period in the hour starting 01:00 to 01:59, 04:00 to 04:59 amd 14:00 to 14:59 (Note excludes report times at 02:00, 05:00 and 15:00).
0 * 1 * 0,3 On the hour on the first of the month and on the hour every Sunday and Wednesday.
* * 21,1-10/3 6 * Every archive period on the 1st, 4th, 7th, 10th and 21st of June.
@monthly Midnight on the 1st of the month.

The wee_reports utility and the report_timing option

The report_timing option is ignored when using the wee_reports utility.

The Standard skin.conf

This section is a reference to the options appearing in the skin configuration file. The default skin is the Standard skin, with a skin configuration file located at SKIN_ROOT/Standard/skin.conf.

The most important options, the ones you are likely to have to customize, are highlighted.

It is worth noting that, like the main configuration file weewx.conf, UTF-8 is used throughout.

[Extras]

This section is available to add any static tags you might want to use in your templates.

Example

As an example, the Standard skin.conf file includes three options:

Skin option Template tag
radar_img $Extras.radar_img
radar_url $Extras.radar_url
googleAnalyticsId $Extras.googleAnalyticsId

If you take a look at the template index.html.tmpl you will see examples of testing for these tags (search the file for the string radar_img to find them).

radar_img

Set to an URL to show a local radar image for you.

radar_url

If the above radar image is clicked, the browser will go to this URL. This is usually used to show a more detailed, close-up, radar picture.

For me in Oregon, setting the above two options to:

[Extras]
    radar_img = http://radar.weather.gov/ridge/lite/N0R/RTX_loop.gif
    radar_url = http://radar.weather.gov/ridge/radar.php?product=NCR&rid=RTX&loop=yes

results in a nice image of a radar centered on Portland, Oregon. When you click on it, it gives you a detailed, animated view. If you live in the USA, take a look at the NOAA radar website to find a nice one that will work for you. In other countries, you will have to consult your local weather service.

googleAnalyticsId

If you have a Google Analytics ID, you can set it here. The Google Analytics Javascript code will then be included, enabling analytics of your website usage. If commented out, the code will not be included.

Extending [Extras]

Other tags can be added in a similar manner, including sub-sections. For example, say you have added a video camera and you would like to add a still image with a hyperlink to a page with the video. You want all of these options to be neatly contained in a sub-section.

[Extras]
    [[video]]
        still = video_capture.jpg
        hyperlink = http://www.eatatjoes.com/video.html
      

Then in your template you could refer to these as:

<a href="$Extras.video.hyperlink">
    <img src="$Extras.video.still" alt="Video capture"/>
</a>

[Units]

This section deals with Units and their formatting.

[[Groups]]

This sub-section lists all the Unit Groups and specifies which unit system is to be used for each one of them.

As there are many different observational measurement types (such as outTemp, barometer, etc.) used in weeWX (more than 50 at last count), it would be tedious, not to say possibly inconsistent, to specify a different measurement system for each one of them. At the other extreme, requiring all of them to be "U.S. Customary" or "Metric" seems overly restrictive. WeeWX has taken a middle route and divided all the different observation types into 12 different unit groups. A unit group is something like group_temperature. It represents the measurement system to be used by all observation types that are measured in temperature, such as inside temperature (type inTemp), outside temperature (outTemp), dewpoint (dewpoint), wind chill (windchill), and so on. If you decide that you want unit group group_temperature to be measured in degree_C then you are saying all members of its group will be reported in degrees Celsius.

Note that the unit system is always specified in the singular. That is, specify degree_C or foot, not degrees_C or feet. See the Appendix: Units for more information, including a concise summary of the groups, their members, and which options can be used for each group.

group_altitude

Which measurement unit to be used for altitude. Possible options are foot or meter.

group_direction

Which measurement unit to be used for direction. The only option is degree_compass.

group_moisture

The measurement unit to be used for soil moisture. The only option is centibar.

group_percent

The measurement unit to be used for percentages. The only option is percent.

group_pressure

The measurement unit to be used for pressure. Possible options are one of inHg (inches of mercury), mbar, or hPa.

group_radiation

The measurement unit to be used for radiation. The only option is watt_per_meter_squared.

group_rain

The measurement unit to be used for precipitation. Options are inch, cm, or mm.

group_rainrate

The measurement unit to be used for rate of precipitation. Possible options are one of inch_per_hour, cm_per_hour, or mm_per_hour.

group_speed

The measurement unit to be used for wind speeds. Possible options are one of mile_per_hour, km_per_hour, knot, or meter_per_second.

group_speed2

This group is similar to group_speed, but is used for calculated wind speeds which typically have a slightly higher resolution. Possible options are one mile_per_hour2, km_per_hour2, knot2, or meter_per_second2.

group_temperature

The measurement unit to be used for temperatures. Options are degree_F or degree_C.

group_volt

The measurement unit to be used for voltages. The only option is volt.

[[StringFormats]]

This sub-section is used to specify what string format is to be used for each unit when a quantity needs to be converted to a string. Typically, this happens with y-axis labeling on plots and for statistics in HTML file generation. For example, the options

degree_C = %.1f
inch     = %.2f

would specify that the given string formats are to be used when formatting any temperature measured in degrees Celsius or any precipitation amount measured in inches, respectively. The formatting codes are those used by Python, and are very similar to C's sprintf() codes.

You can also specify what string to use for an invalid or unavailable measurement (value None). For example,

NONE = " N/A "

[[Labels]]

This sub-section specifies what label is to be used for each measurement unit type. For example, the options

degree_F = °F
inch     = ' in'

would cause all temperatures to have unit labels °F and all precipitation to have labels in. If any special symbols are to be used (such as the degree sign above) they should be encoded in UTF-8. This is generally what most text editors use if you cut-and-paste from a character map.

If the label includes two values, then the first is assumed to be the singular form, the second the plural form. For example,

foot   = " foot",   " feet"
...
day    = " day",    " days"
hour   = " hour",   " hours"
minute = " minute", " minutes"
second = " second", " seconds"

This is particularly useful when localizing the weeWX and server uptimes.

[[TimeFormats]]

This sub-section is used for time labels. It uses strftime() formats. The default looks like this:

    [[TimeFormats]]
        # This section sets the string format to be used for each time scale.
        # The values below will work in every locale, but may not look
        # particularly attractive. See the Customization Guide for alternatives.

        day        = %X
        week       = %X (%A)
        month      = %x %X
        year       = %x %X
        rainyear   = %x %X
        current    = %x %X
        ephem_day  = %X
        ephem_year = %x %X

The specifiers %x, %X, and %A code locale dependent date, time, and weekday names, respectively. Hence, if you set an appropriate environment variable LANG, then the date and times should follow local conventions (see section Environment variable LANG for details on how to do this). However, they will not look particularly pretty and you may want to change them. For example, I use this in the U.S.:

    [[TimeFormats]]
        #
        # More attractive formats that work in most Western countries.
        #
        day        = %H:%M
        week       = %H:%M on %A
        month      = %d-%b-%Y %H:%M
        year       = %d-%b-%Y %H:%M
        rainyear   = %d-%b-%Y %H:%M
        current    = %d-%b-%Y %H:%M
        ephem_day  = %H:%M
        ephem_year = %d-%b-%Y %H:%M

The last two formats, ephem_day and ephem_year allow the formatting to be set for almanac times The first, ephem_day, is used for almanac times within the day, such as sunrise or sunset. The second, ephem_year, is used for almanac times within the year, such as the next equinox or full moon.

[[Ordinates]]

directions

Set to the abbreviations to be used for ordinal directions. By default, this is N, NNE, NE, ENE, E, ESE, SE, SSE, S, SSW, SW, WSW, W, WNW, NW, NNW, N.

[[DegreeDays]]

heating_base
cooling_base

Set to the base temperature for calculating heating and cooling degree-days, along with the unit to be used. Examples:

heating_base = 65.0, degree_F
cooling_base = 20.0, degree_C

[[Trend]]

time_delta

Set to the time difference over which you want trends to be calculated. The default is 3 hours.

time_grace

When searching for a previous record to be used in calculating a trend, a record within this amount of time_delta will be accepted. Default is 300 seconds.

[Labels]

This section sets the various labels to use.

hemispheres

Comma separated list for the labels to be used for the four hemispheres. The default is N, S, E, W.

latlon_formats

Comma separated list for the formatting to be used when converting latitude and longitude to strings. There should be three elements:

  1. The format to be used for whole degrees of latitude
  2. The format to be used for whole degrees of longitude
  3. The format to be used for minutes.

This allows you to decide whether or not you want leading zeroes. The default includes leading zeroes and is "%02d", "%03d", "%05.2f"

[[Generic]]

This sub-section specifies default labels to be used for each observation type. For example, options

inTemp  = Temperature inside the house
outTemp = Outside Temperature
UV      = UV Index

would cause the given labels to be used for plots of inTemp and outTemp. If no option is given, then the observation type itself will be used (e.g., outTemp).

[Almanac]

This section controls what text to use for the almanac. It consists of only one entry

moon_phases

This option is a comma separated list of labels to be used for the eight phases of the moon. Default is New, Waxing crescent, First quarter, Waxing gibbous, Full, Waning gibbous, Last quarter, Waning crescent.

[CheetahGenerator]

This section is used by generator weewx.cheetahgenerator.CheetahGenerator and controls text generation from templates, specifically which files are to be produced from which template.

Overview of file generation

Files are generated from templates, and each template is identified by the template parameter.

Each template file is named something like D/F.E.tmpl, where D is the (optional) directory the template sits in and will also be the directory the results will be put in, and F.E is the generated file name. So, given a template file with name Acme/index.html.tmpl, the results will be put in HTML_ROOT/Acme/index.html.

The configuration for a group of templates will look something like this:

[CheetahGenerator]
    [[index]]
        template = index.html.tmpl
    [[textfile]]
        template = filename.txt.tmpl
    [[xmlfile]]
        template = filename.xml.tmpl

There can be only one template in each block. In most cases, the block name does not matter — it is used only to isolate each template. However, there are three block names that have special meaning: SummaryByMonth, SummaryByYear, and ToDate. They are described below.

The file generator runs on each new archive record. In a default weeWX installation, that would be every 5 minutes.

Cheetah processes each template to generate a file. Cheetah follows any logic defined by directives such as for or if ... else, and it replaces variables such as $Extras.radar_url or $current.outTemp.max.

Variables are defined by objects in weeWX. Some variables are static, others are linked to data in databases. The list of variables can be extended.

File generation options

search_list

This is the list of search list objects that will be scanned by the template engine, looking for tags. See the section Defining new tags and the Cheetah documentation for details on search lists. If no search_list is specified, a default list will be used. The default list is:

search_list = weewx.cheetahgenerator.Almanac, weewx.cheetahgenerator.Station, weewx.cheetahgenerator.Stats, weewx.cheetahgenerator.UnitInfo, weewx.cheetahgenerator.Extras

search_list_extensions

This defines one or more search list objects that will be appended to the search_list. For example, the following adds alltime and forecast variables to the search list.

search_list_extensions = user.stats.MyStats, user.forecast.ForecastVariables

encoding

As Cheetah goes through the template, it substitutes strings for all tag values. This option controls which encoding to use for the new strings. The encoding can be chosen on a per file basis. There are 3 possible choices:

Encoding Comments
html_entities Non 7-bit characters will be represented as HTML entities (e.g., the degree sign will be represented as &#176;)
utf8 Non 7-bit characters will be represented in UTF-8.
strict_ascii Non 7-bit characters will be ignored.

The encoding html_entities is the default.

template

The name of a template file. A template filename must end with .tmpl. Filenames are case-sensitive. If the template filename has the letters YYYY or MM in its name, these will be substituted for the year and month, respectively. So, a template with the name summary-YYYY-MM.html.tmpl would have name summary-2010-03.html for the month of March, 2010.

stale_age

File staleness age, in seconds. If the file is older than this age it will be generated from the template. If no stale_age is specified, then the file will be generated every time the generator runs.

[[SummaryByMonth]]

The SummaryByMonth section defines some special behavior. Each template in this section will be used multiple times, each time with a different per-month timespan. Be sure to include YYYY and MM in the filename of any template in this section.

[[SummaryByYear]]

The SummaryByYear section defines some special behavior. Each template in this section will be used multiple times, each time with a different per-year timespan. Be sure to include YYYY in the filename of any template in this section.

Customizing file generation

The best way to customize file generation is to make a copy of a working report/skin, then make incremental changes.

When there is an error during template generation, the error will show up in the log file. Many errors are obvious — Cheetah will display a line number and list the template file in which the error occurred. In some cases the error reporting is rather obscure. So make small changes and test often. Use the utility wee_reports to test modifications to the generator configuration and/or the template contents.

The Standard skin templates

Here is the [CheetahGenerator] section from the Standard skin.conf

[CheetahGenerator]
    # This section is used by the generator CheetahGenerator, and specifies
    # which files are to be generated from which template.

    encoding = html_entities

    [[SummaryByMonth]]
        # Reports that summarize "by month"
        [[[NOAA_month]]]
            encoding = strict_ascii
            template = NOAA/NOAA-YYYY-MM.txt.tmpl

    [[SummaryByYear]]
        # Reports that summarize "by year"
        [[[NOAA_year]]]
            encoding = strict_ascii
            template = NOAA/NOAA-YYYY.txt.tmpl

    [[ToDate]]
        # Reports that show statistics "to date", such as day-to-date,
        # week-to-date, month-to-date, etc.
        [[[day]]]
            template = index.html.tmpl

        [[[week]]]
            template = week.html.tmpl

        [[[month]]]
            template = month.html.tmpl

        [[[year]]]
            template = year.html.tmpl

        [[[RSS]]]
            template = RSS/weewx_rss.xml.tmpl

        [[[Mobile]]]
            template = mobile.html.tmpl

The Standard skin contains three different kinds of generated output:

  1. Summary by Month. The Standard skin uses SummaryByMonth to produce NOAA summaries, one for each month, as a simple text file.
  2. Summary by Year. The Standard skin uses SummaryByYear to produce NOAA summaries, one for each year, as a simple text file.
  3. Summary "To Date". The Standard skin produce reports for the day, week, month, and year-to-date observations. These files are HTML. The first, the daily summary (output filename is index.html), includes a drop-down list that displays the NOAA month and yearly summaries.

The encoding for text files is strict_ascii, whereas the encoding for html files is html_entities. In the Standard skin this is specified by declaring encoding = html_entities at the top level of [CheetahGenerator] then encoding = strict_ascii for each text file.

Other than SummaryByMonth and SummaryByYear, the section names are arbitrary. ToDate could just as well have been called files_to_date, and the sections day, week, and month could just as well have been called tom, dick, and harry.

[ImageGenerator]

This section describes the various options available to the image generator.

Part names in a weeWX image
Part names in a weeWX image

Overall options

These are options that affect the overall image.

image_width
image_height

The width and height of the image in pixels. Optional. Default is 300 x 180 pixels.

image_background_color

The background color of the whole image. Optional. Default is 0xf5f5f5 ("SmokeGray")

chart_background_color

The background color of the chart itself. Optional. Default is 0xd8d8d8.

chart_gridline_color

The color of the chart grid lines. Optional. Default is 0xa0a0a0

Effect of anti_alias option
A GIF showing the same image with anti_alias=1, 2, and 4.
Example of day/night bands
Example of day/night bands in a one week image

anti_alias

Setting to 2 or more might give a sharper image, with fewer jagged edges. Experimentation is in order. Default is 1.

show_daynight

Set to true to show day/night bands in an image. Otherwise, set to false. This only looks good with day or week plots. Optional. Default is false.

daynight_day_color

The color to be used for the daylight band. Optional. Default is 0xffffff.

daynight_night_color

The color to be used for the nighttime band. Optional. Default is 0xf0f0f0, a dark gray.

daynight_edge_color

The color to be used in the transition zone between night and day. Optional. Default is 0xefefef, a mid-gray.

Various label options

These are options for the various labels used in the image.

top_label_font_path

The path to the font to be use for the top label. Optional. If not given, or if weeWX cannot find the font, then the default PIL font will be used.

top_label_font_size

The size of the top label in pixels. Optional. The default is 10.

unit_label_font_path

The path to the font to be use for the unit label. Optional. If not given, or if weeWX cannot find the font, then the default PIL font will be used.

unit_label_font_size

The size of the unit label in pixels. Optional. The default is 10.

unit_label_font_color

The color of the unit label font. Optional. Default is black.

bottom_label_font_path

The path to the font to be use for the bottom label. Optional. If not given, or if weeWX cannot find the font, then the default PIL font will be used.

bottom_label_font_size

The size of the bottom label in pixels. Optional. The default is 10.

bottom_label_font_color

The color of the bottom label font. Optional. Default is black.

bottom_label_format

The format to be used for the bottom label. It should be a strftime format. Optional. Default is '%m/%d/%y %H:%M'.

bottom_label_offset

The margin of the bottom label from the bottom of the plot. Default is 3.

axis_label_font_path

The path to the font to be use for the x- and y-axis labels. Optional. If not given, or if weeWX cannot find the font, then the default PIL font will be used.

axis_label_font_size

The size of the x- and y-axis labels in pixels. Optional. The default is 10.

axis_label_font_color

The color of the x- and y-axis label font. Optional. Default is black.

x_label_format

The format to be used for the time labels on the x-axis. It should be a strftime format. Optional. If not given, a sensible format will be chosen automatically.

Plot scaling options

time_length

The nominal length of the time period to be covered in seconds. The exact length of the x-axis is chosen by the plotting engine to cover this period. Optional. Default is 86400 (one day).

yscale

A 3-way tuple (ylow, yhigh, min_interval), where ylow and yhigh are the minimum and maximum y-axis values, respectively, and min_interval is the minimum tick interval. If set to None, the corresponding value will be automatically chosen. Optional. Default is None, None, None. (Choose the y-axis minimum, maximum, and minimum increment automatically.)

Compass rose options

Example of a progressive vector plot
Example of a vector plot with a compass rose in the lower-left

rose_label

The label to be used in the compass rose to indicate due North. Optional. Default is N.

rose_label_font_path

The path to the font to be use for the rose label (the letter "N," indicating North). Optional. If not given, or if weeWX cannot find the font, then the default PIL font will be used.

rose_label_font_size

The size of the compass rose label in pixels. Optional. The default is 10.

rose_label_font_color

The color of the compass rose label. Optional. Default is the same color as the rose itself.

vector_rotate

Causes the vectors to be rotated by this many degrees. Positive is clockwise. If westerly winds dominate at your location (as they do at mine), then you may want to specify +90 for this option. This will cause the average vector to point straight up, rather than lie flat against the x-axis. Optional. The default is 0.

Shared plot line options

These are options shared by all the plot lines.

chart_line_colors

Each chart line is drawn in a different color. This option is a list of those colors. If the number of lines exceeds the length of the list, then the colors wrap around to the beginning of the list. NB: individual line color can be overridden by using option color. Optional. In the case of bar charts, this is the color of the outline of the bar. Default is 0xff0000, 0x00ff00, 0x0000ff.

chart_fill_colors

A list of the color to be used as the fill of the bar charts. Optional. The default is to use the same color as the outline color (option chart_line_colors) above.

chart_line_width

Each chart line can be drawn using a different line width. This option is a list of these widths. If the number of lines exceeds the length of the list, then the widths wrap around to the beginning of the list. NB: individual line widths can be overridden by using option width. Optional. Default is 1, 1, 1.

Individual line options

These are options that are set for individual lines.

aggregate_type

The default is to plot every data point, but this is probably not a good idea for any plot longer than a day. By setting this option, you can aggregate data by a set time interval. Available aggregation types include avg, sum, max, min, count, and last.

aggregate_interval

The time period over which the data should be aggregated. Required if aggregate_type has been set.

plot_type

The type of plot for this line. Choices are line, bar, or vector. Optional. Default is line.

color

This option is to override the color for an individual line. Optional. Default is to use the color in chart_line_colors.

line_type

The type of line to be used. Choices are solid or none. Optional. Default is solid.

marker_type

The type of marker to be used to marke each data point. Choices are cross, x, circle, box, or none. Optional. Default is none.

marker_size

The size of the marker. Optional. Default is 8.

line_gap_fraction

If there is a gap between data points bigger than this fractional amount of the x-axis, then a gap will be drawn, rather than a connecting line. See Section Line gaps. Optional. The default is to always draw the line.

width

This option is to override the line widthfor an individual line. Optional. Default is to use the width in chart_line_width.

label

The label to be used for this plot line in the top label. Optional. The default is to use the SQL variable name.

data_type

The SQL data type to be used for this plot line. For more information, see the section Including the same SQL type more than once in a plot. Optional. The default is to use the section name.

[CopyGenerator]

This section is used by generator weewx.reportengine.CopyGenerator and controls which files are to be copied over from the skin directory to the destination directory. Think of it as "file generation," except that rather than going through the template engine, the files are simply copied over.

copy_once

This option controls which files get copied over on the first invocation of the report engine service. Typically, this is things such as style sheets or background GIFs. Wildcards can be used.

copy_always

This is a list of files that should be copied on every invocation. Wildcards can be used.

Here is the [CopyGenerator] section from the Standard skin.conf

[CopyGenerator]
    # This section is used by the generator CopyGenerator

    # List of files to be copied only the first time the generator runs
    copy_once = backgrounds/*, weewx.css, mobile.css, favicon.ico

    # List of files to be copied each time the generator runs
    # copy_always = 

The Standard skin includes some background images, CSS files, and icons that need to be copied once. There are no files that need to be copied every time the generator runs.

[Generators]

This section defines the list of generators that should be run.

generator_list

This option controls which generators get run for this skin. It is a comma separated list. The generators will be run in this order.

Here is the [Generators] section from the Standard skin.conf

[Generators]
    generator_list = weewx.cheetahgenerator.CheetahGenerator, weewx.imagegenerator.ImageGenerator, weewx.reportengine.CopyGenerator

The Standard skin uses three generators: CheetahGenerator, ImageGenerator, and CopyGenerator.

Localization

What follows is a guide to localizing to a non-English language and/or locale. There are two parts: translating to different languages and modifying to reflect local conventions for displaying data.

Translate the templates

First, you will need to go through the templates and translate to your target language. Obvious text strings such as "Current Weather Conditions" will need to be translated.

Modify the skin configuration

Next, you will need to go through skin.conf to translate labels and modify formats to follow local conventions.

You will probably want to change the generic labels used for the observation types:

[Labels]
    ...
    [[Generic]]
        barometer      = Barometer
        dewpoint       = Dew Point
        heatindex      = Heat Index
        inHumidity     = Inside Humidity
        inTemp         = Inside Temperature
        outHumidity    = Outside Humidity
        outTemp        = Outside Temperature
        radiation      = Radiation
        rain           = Rain
        rainRate       = Rain Rate
        rxCheckPercent = ISS Signal Quality
        windDir        = Wind Direction
        windGust       = Gust Speed
        windGustDir    = Gust Direction
        windSpeed      = Wind Speed
        windchill      = Wind Chill
        windgustvec    = Gust Vector
        windvec        = Wind Vector

The hemisphere abbreviations may have to be changed:

[Labels]
    hemispheres = N, S, E, W

The wind ordinal directions may have to be changed:

[Units]
    ...
    [[Ordinates]]
        # The ordinal directions. The last one should be for no wind direction
        directions = N, NNE, NE, ENE, E, ESE, SE, SSE, S, SSW, SW, WSW, W, WNW, NW, NNW, N/A 

Don't forget the moon phases:

[Almanac]
    moon_phases = New, Waxing crescent, First quarter, Waxing gibbous, Full, Waning gibbous, Last quarter, Waning crescent

Most of the unit labels either follow ISO conventions, or are unlikely to be used outside English speaking countries (an example would be "foot"). But, there are a few exceptions, used to label the weeWX and server "uptimes", which can be found in sub-section [Units][[Labels]]

[Units]
    ...
    [[Labels]]
        ...
        day     = " day",    " days"
        hour    = " hour",   " hours"
        minute  = " minute", " minutes"
        second  = " second", " seconds"

By default, the time formats will use the local convention specified by the LANG environment variable. These can be modified individually in the [Units][[TimeFormats]] section.

[Units]
    [[TimeFormats]]
        day        = %X
        week       = %X (%A)
        month      = %x %X
        year       = %x %X
        rainyear   = %x %X
        current    = %x %X
        ephem_day  = %X
        ephem_year = %x %X

Encodings

You may have to specify the encoding used

Encoding used in a template

If you use an encoding other than ASCII in a template, then you must include a #encoding directive at the top of the file, informing Cheetah of your choice. If your template includes other templates (by using the #include directive), and those templates use an encoding, then they must include their own #encoding directive as well.

For example, if you use UTF-8 (the default used by all the standard weeWX templates), then the top of your template should look something like this:

#errorCatcher Echo
##
## Specifying an encoding of UTF-8 is usually safe, but if your text is
## actually in Latin-1, then you should replace the string "UTF-8" with "latin-1"
#encoding UTF-8

<!DOCTYPE HTML PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">

Encoding used in substituted strings

The other place where encoding matters is in the strings substituted by Cheetah. The encoding for these is specified by the encoding option in the skin configuration file, skin.conf.

Environment variable LANG

Finally, you will need to set the environment variable LANG to reflect your locale. For example, assuming you set

$ export LANG=es_ES.UTF-8

before running weeWX, then the local Spanish names for days of the week and months of the year will be used. The decimal point for numbers will also be modified appropriately.

Customizing the weeWX service engine

This is an advanced topic intended for those who wish to try their hand at extending the internal engine in weeWX. Before attempting these examples, you should be reasonably proficient with Python.

Please note that the API to the service engine may change in future versions!

At a high level, weeWX consists of an engine that is responsible for managing a set of services. A service consists of a Python class which binds its member functions to various events. The engine arranges to have the bound member function called when a specific event happens, such as a new LOOP packet arriving.

To customize, you can

See the table Default services above for a list of the services that are normally run.

Modifying an existing service

The service weewx.engine.StdPrint prints out new LOOP and archive packets to the console when they arrive. By default, it prints out the entire record, which generally includes a lot of possibly distracting information and can be rather messy. Suppose you do not like this, and want it to print out only the time, barometer reading, and the outside temperature whenever a new LOOP packet arrives. This could be done by subclassing the default print service StdPrint and overriding member function new_loop_packet().

Create the file user/myprint.py:

from weewx.engine import StdPrint
from weeutil.weeutil import timestamp_to_string

class MyPrint(StdPrint):

    # Override the default new_loop_packet member function:
    def new_loop_packet(self, event):
        packet = event.packet
        print "LOOP: ", timestamp_to_string(packet['dateTime']),
            "BAR=",  packet.get('barometer', 'N/A'),
            "TEMP=", packet.get('outTemp', 'N/A')

This service substitutes a new implementation for the member function new_loop_packet. This implementation prints out the time, then the barometer reading (or N/A if it is not available) and the outside temperature (or N/A).

You then need to specify that your print service class should be loaded instead of the default StdPrint service. This is done by substituting your service name for StdPrint in service_list, located in [Engine][[Services]]:

[Engine]
    [[Services]]
        ...
        report_services = user.myprint.MyPrint, weewx.engine.StdReport

Note that the report_services must be all on one line. Unfortunately, the parser ConfigObj does not allow options to be continued on to following lines.

Creating a new service

Suppose there is no service that can be easily customized for your needs. In this case, a new one can easily be created by subclassing off the abstract base class StdService, and then adding the functionality you need. Here is an example that implements an alarm, which sends off an email when an arbitrary expression evaluates True.

This example is included in the standard distribution as examples/alarm.py:

import time
import smtplib
from email.mime.text import MIMEText
import threading
import syslog

import weewx
from weewx.engine import StdService
from weeutil.weeutil import timestamp_to_string, option_as_list

# Inherit from the base class StdService:
class MyAlarm(StdService):
    """Service that sends email if an arbitrary expression evaluates true"""
    
    def __init__(self, engine, config_dict):
        # Pass the initialization information on to my superclass:
        super(MyAlarm, self).__init__(engine, config_dict)
        
        # This will hold the time when the last alarm message went out:
        self.last_msg_ts = 0
        
        try:
            # Dig the needed options out of the configuration dictionary.
            # If a critical option is missing, an exception will be raised and
            # the alarm will not be set.
            self.expression    = config_dict['Alarm']['expression']
            self.time_wait     = int(config_dict['Alarm'].get('time_wait', 3600))
            self.smtp_host     = config_dict['Alarm']['smtp_host']
            self.smtp_user     = config_dict['Alarm'].get('smtp_user')
            self.smtp_password = config_dict['Alarm'].get('smtp_password')
            self.SUBJECT       = config_dict['Alarm'].get('subject', "Alarm message from weeWX")
            self.FROM          = config_dict['Alarm'].get('from', 'alarm@example.com')
            self.TO            = option_as_list(config_dict['Alarm']['mailto'])
            syslog.syslog(syslog.LOG_INFO, "alarm: Alarm set for expression: '%s'" % self.expression)
            
            # If we got this far, it's ok to start intercepting events:
            self.bind(weewx.NEW_ARCHIVE_RECORD, self.newArchiveRecord)    # NOTE 1
        except KeyError, e:
            syslog.syslog(syslog.LOG_INFO, "alarm: No alarm set.  Missing parameter: %s" % e)
            
    def newArchiveRecord(self, event):
        """Gets called on a new archive record event."""
        
        # To avoid a flood of nearly identical emails, this will do
        # the check only if we have never sent an email, or if we haven't
        # sent one in the last self.time_wait seconds:
        if not self.last_msg_ts or abs(time.time() - self.last_msg_ts) >= self.time_wait :
            # Get the new archive record:
            record = event.record
            
            # Be prepared to catch an exception in the case that the expression contains 
            # a variable that is not in the record:
            try:                                                              # NOTE 2
                # Evaluate the expression in the context of the event archive record.
                # Sound the alarm if it evaluates true:
                if eval(self.expression, None, record):                       # NOTE 3
                    # Sound the alarm!
                    # Launch in a separate thread so it doesn't block the main LOOP thread:
                    t  = threading.Thread(target = MyAlarm.soundTheAlarm, args=(self, record))
                    t.start()
                    # Record when the message went out:
                    self.last_msg_ts = time.time()
            except NameError, e:
                # The record was missing a named variable. Write a debug message, then keep going
                syslog.syslog(syslog.LOG_DEBUG, "alarm: %s" % e)

    def soundTheAlarm(self, rec):
        """This function is called when the given expression evaluates True."""
        
        # Get the time and convert to a string:
        t_str = timestamp_to_string(rec['dateTime'])

        # Log it
        syslog.syslog(syslog.LOG_INFO, "alarm: Alarm expression \"%s\" evaluated True at %s" % (self.expression, t_str))

        # Form the message text:
        msg_text = "Alarm expression \"%s\" evaluated True at %s\nRecord:\n%s" % (self.expression, t_str, str(rec))
        # Convert to MIME:
        msg = MIMEText(msg_text)
        
        # Fill in MIME headers:
        msg['Subject'] = self.SUBJECT
        msg['From']    = self.FROM
        msg['To']      = ','.join(self.TO)
        
        # Create an instance of class SMTP for the given SMTP host:
        s = smtplib.SMTP(self.smtp_host)
        try:
            # Some servers (eg, gmail) require encrypted transport.
            # Be prepared to catch an exception if the server
            # doesn't support it.
            s.ehlo()
            s.starttls()
            s.ehlo()
            syslog.syslog(syslog.LOG_DEBUG, "alarm: using encrypted transport")
        except smtplib.SMTPException:
            syslog.syslog(syslog.LOG_DEBUG, "alarm: using unencrypted transport")

        try:
            # If a username has been given, assume that login is required for this host:
            if self.smtp_user:
                s.login(self.smtp_user, self.smtp_password)
                syslog.syslog(syslog.LOG_DEBUG, "alarm: logged in with user name %s" % (self.smtp_user,))
            
            # Send the email:
            s.sendmail(msg['From'], self.TO,  msg.as_string())
            # Log out of the server:
            s.quit()
        except Exception, e:
            syslog.syslog(syslog.LOG_ERR, "alarm: SMTP mailer refused message with error %s" % (e,))
            raise
        
        # Log sending the email:
        syslog.syslog(syslog.LOG_INFO, "alarm: email sent to: %s" % self.TO)

This service expects all the information it needs to be in the configuration file weewx.conf in a new section called [Alarm]. So, add the following lines to your configuration file:

[Alarm]
    expression = "outTemp < 40.0"
    time_wait = 3600
    smtp_host = smtp.example.com
    smtp_user = myusername
    smtp_password = mypassword
    mailto = auser@example.com, anotheruser@example.com
    from   = me@example.com
    subject = "Alarm message from weeWX!"

There are three important points to be noted in this example, each marked with a NOTE flag in the code.

  1. Here is where the binding happens between an event, weewx.NEW_ARCHIVE_RECORD in this example, and a member function, self.newArchiveRecord. When the event NEW_ARCHIVE_RECORD occurs, the function self.newArchiveRecord will be called. There are many other events that can be intercepted. Look in the file weewx/__init__.py.
  2. Some hardware does not emit all possible observation types in every record. So, it's possible that a record may be missing some types that are used in the expression. This try block will catch the NameError exception that would be raised should this occur.
  3. This is where the test is done for whether or not to sound the alarm. The [Alarm] configuration options specify that the alarm be sounded when outTemp < 40.0 evaluates True, that is when the outside temperature is below 40.0 degrees. Any valid Python expression can be used, although the only variables available are those in the current archive record.

Another example expression could be:

expression = "outTemp < 32.0 and windSpeed > 10.0"

In this case, the alarm is sounded if the outside temperature drops below freezing and the wind speed is greater than 10.0.

Note that units must be the same as whatever is being used in your database. That is, the same as what you specified in option target_unit.

Option time_wait is used to avoid a flood of nearly identical emails. The new service will wait this long before sending another email out.

Email will be sent through the SMTP host specified by option smtp_host. The recipient(s) are specified by the comma separated option mailto.

Many SMTP hosts require user login. If this is the case, the user and password are specified with options smtp_user and smtp_password, respectively.

The last two options, from and subject are optional. If not supplied, weeWX will supply something sensible. Note, however, that some mailers require a valid "from" email address and the one weeWX supplies may not satisfy its requirements.

To make this all work, you must first copy the alarm.py file to the user directory. Then tell the engine to load this new service by adding the service name to the list report_services, located in [Engine][[Services]]:

[Engine]
   [[Services]]
        report_services = weewx.engine.StdPrint, weewx.engine.StdReport, user.alarm.MyAlarm

Again, note that the option report_services must be all on one line — the parser ConfigObj does not allow options to be continued on to following lines.

In addition to the example above, the distribution also includes a low-battery alarm (lowBattery.py), which is similar, except that it intercepts LOOP events (instead of archiving events).

Adding a second data source

A very common problem is wanting to augment the data from your weather station with data from some other device. Generally, you have two approaches for how to handle this:

This section covers the latter approach.

Suppose you have installed an electric meter at your house and you wish to correlate electrical usage with the weather. The meter has some sort of connection to your computer, allowing you to download the total power consumed. At the end of every archive interval you want to calculate the amount of power consumed during the interval, then add the results to the record coming off your weather station. How would you do this?

Here is the outline of a service that retrieves the electrical consumption data and adds it to the archive record. It assumes that you already have a function download_total_power() that, somehow, downloads the amount of power consumed since time zero.

File user/electricity.py

import weewx
from weewx.engine import StdService

class AddElectricity(StdService):

    def __init__(self, engine, config_dict):

      # Initialize my superclass first:
      super(AddElectricity, self).__init__(engine, config_dict)

      # Bind to any new archive record events:
      self.bind(weewx.NEW_ARCHIVE_RECORD, self.new_archive_packet)
      
      self.last_total = None

    def new_archive_packet(self, event):

        total_power = download_total_power()
        
        if self.last_total:
            net_consumed = total_power - self.last_total
            event.record['electricity'] = net_consumed

        self.last_total = total_power

This adds a new key electricity to the record dictionary and sets it equal to the difference between the amount of power currently consumed and the amount consumed at the last archive record. Hence, it will be the amount of power consumed over the archive interval. The unit should be Watt-hours.

As an aside, it is important that the function download_total_power() does not delay very long because it will sit right in the main loop of the weeWX engine. If it's going to cause a delay of more than a couple seconds you might want to put it in a separate thread and feed the results to AddElectricity through a queue.

To make sure your service gets run, you need to add it to one of the service lists in weewx.conf, section [Engine], subsection [[Services]]. This subsection lists all the services to be run, broken up into different service lists. Here's a description of them:

Service list Function
prep_services Perform any actions before the main loop is run.
data_services Augment data, before it is processed.
process_services Process, filter, and massage the data.
archive_services Record the data in a database.
restful_services Upload processed data to an external RESTful service.
report_services Run any reports.

These lists are designed to orchestrate the data as it flows through the weeWX engine. For example, you want to make sure data has been processed by, for example, running it through the quality control service, StdQC, before putting them in the database. Similarly, the reporting system must come after the archiving service. These groups insure that things happen in the proper sequence.

In our case, the obvious place for our new service is in data_services. When you're done, your section [Engine] will look something like this:

#   This section configures the internal weeWX engine.

[Engine]
    
    [[Services]]
        # This section specifies the services that should be run. They are
        # grouped by type, and the order of services within each group
        # determines the order in which the services will be run.
        prep_services = weewx.engine.StdTimeSynch
        data_services = user.electricity.AddElectricity
        process_services = weewx.engine.StdConvert, weewx.engine.StdCalibrate, weewx.engine.StdQC, weewx.wxservices.StdWXCalculate
        archive_services = weewx.engine.StdArchive
        restful_services = weewx.restx.StdStationRegistry, weewx.restx.StdWunderground, weewx.restx.StdPWSweather, weewx.restx.StdCWOP, weewx.restx.StdWOW, weewx.restx.StdAWEKAS
        report_services = weewx.engine.StdPrint, weewx.engine.StdReport

Customizing the database

For most users the default database will work just fine. It has the added advantage of being compatible with the wview database. Nevertheless, there may be occasions where you may want to add a new SQL type to your database, or change its unit system. This section shows you how to do this.

Add a new type to the archive database

If you followed the example above, Adding a second data source, you created a new observation type, electricity. Trouble is, there is no corresponding type in the schema of the SQL database and, therefore, weeWX won't know where to store it. How would you add such a type?

The utility wee_database can be used to do this. It will create a new database that is similar to the old database, except it will have the new type in its schema.

Here's our general strategy:

  1. Extend the existing schema with the new type electricity.
  2. Modify the wx_binding to use the new schema instead of the old.
  3. Make sure you have the necessary permissions to create a new database.
  4. Use the utility wee_database to create the new database and populate it with data from the old database.
  5. Shuffle databases around so weeWX will use the new database.

Here is the recipe that follows that strategy:

  1. Add the new type to our existing schema. The weather schema that comes with weeWX is the same as what wview uses. It's located in the file schemas/wview.py — take a look at it now. We could just modify it in situ, but that would run the risk of confusing the two versions. Alternatively, we could copy the file over to a new location, and then modify that. But, because our change is just a simple addition, we're going to import the wview schema and just add on our new type. There's also no reason to create a new file. Why not just do it in the file we already have, user/electricity.py? Add to the bottom of the file:

    import schemas.wview
    schema_with_electricity = schemas.wview.schema + [('electricity', 'REAL')]

    This creates a new schema (it will have the name user.electricity.schema_with_electricity), that is just like the old one, except it has a new type electricity tacked on to the end.

  2. Modify wx_binding. When it creates the new, modified database, wee_database needs to know to use your new schema instead of the old, wview schema. You do this by changing the option schema in section [DataBindings] in weewx.conf. It will look like this when you're done:

    [DataBindings]
        [[wx_binding]]
            database = archive_sqlite
            table_name = archive
            manager = weewx.wxmanager.WXDaySummaryManager
            schema = user.electricity.schema_with_electricity
  3. Check permissions. wee_database will create a new database with the same name as the old, except with the suffix _new attached to the end. Make sure you have the necessary permissions to create it. In particular, if you are using MySQL, you will need CREATE privileges:

    mysql> GRANT select, update, create, delete, insert ON weewx_new.* TO weewx@localhost;
  4. Create and populate the new database. Now run the utility wee_database with the --reconfigure option and the path to the configuration file:

    wee_database weewx.conf --reconfigure

    This will create a new database (nominally, weewx.sdb_new if you are using SQLite, weewx_new if you are using MySQL) using the new schema and populate it with data from the old database.

  5. Shuffle the databases. Now arrange things so weeWX can find the new database.

    Warning!
    Make a backup of the data before doing any of the next steps!

    You can either shuffle the databases around so the new database has the same name as the old database, or edit weewx.conf to use the new database name. To do the former:

    For SQLite:

    cd SQLITE_ROOT
    mv weewx.sdb_new weewx.sdb

    For MySQL:

    mysql -u <username> --password=<mypassword>
    mysql> DROP DATABASE weewx;                             # Delete the old database
    mysql> CREATE DATABASE weewx;                           # Create a new one with the same name
    mysql> RENAME TABLE weewx_new.archive TO weewx.archive; # Rename to the nominal name
  6. It's worth noting that there's actually a hidden, last step: rebuilding the daily summaries inside the new database. This will be done automatically by weeWX at the next startup. Alternatively, it can be done manually using the wee_database utility and the --rebuild-daily option:

    wee_database weewx.conf --rebuild-daily

Changing the unit system

Normally, data are stored in the databases using US Customary units and, normally, you don't care; it is an "implementation detail". Data can always be displayed using any units you want. The Customizing reports section explains how to do this. Nevertheless, there may be special situations where you wish to store the data in Metric units. For example, you may need to allow direct programmatic access to the databases from another piece of software that expects metric units.

WeeWX does not allow you to change the database unit system midstream. You can't start with one unit system then, in the middle of the database, switch to another. See the section [StdConvert] in the weeWX User's Guide. However, you can reconfigure the database by copying it to a new database, performing the unit conversion along the way. You then use this new database.

Here is the general strategy:

  1. Modify weewx.conf to reflect your choice of the new unit system to use. Your choices are US, METRIC, or METRICWX. See the Appendix: Units for the exact differences between these three choices.
  2. Make sure you have the necessary permissions to create the new database.
  3. Use the utility wee_database to create the new database and populate it with data from the old database.
  4. Shuffle databases around so weeWX will use the new database.

Here is the recipe that follows that strategy:

  1. Modify weewx.conf. Edit the configuration file to change option target_unit in section [StdConvert] to reflect your choice. If you are switching to metric units, the option will look like:

    [StdConvert]
        target_unit = METRICWX
  2. Check permissions. The reconfiguration utility will create a new database with the same name as the old, except with the suffix _new attached to the end. Make sure you have the necessary permissions to do this. In particular, if you are using MySQL, you will need CREATE privileges.

  3. Create and populate the new database. Now run the utility wee_database with the --reconfigure option:

    wee_database weewx.conf --reconfigure

    This will create a new database (nominally, weewx.sdb_new if you are using SQLite, weewx_new if you are using MySQL), using the schema found in user/schemas.py, and populate it with data from the old database, while performing the unit conversion.

  4. Shuffle the databases. Now arrange things so weeWX can find the new database.

    Warning!
    Make a backup of the data before doing any of the next steps!

    You can either shuffle the databases around so the new database has the same name as the old database, or edit weewx.conf to use the new database name. To do the former:

    For SQLite:

    cd SQLITE_ROOT
    mv weewx.sdb_new weewx.sdb

    For MySQL:

    mysql -u <username> --password=<mypassword>
    mysql> DROP DATABASE weewx;                             # Delete the old database
    mysql> CREATE DATABASE weewx;                           # Create a new one with the same name
    mysql> RENAME TABLE weewx_new.archive TO weewx.archive; # Rename to the nominal name
  5. It's worth noting that there's actually a hidden, last step: rebuilding the daily summaries inside the new database. This will be done automatically by weeWX at the next startup. Alternatively, it can be done manually using the wee_database utility and the --rebuild-daily option:

    wee_database weewx.conf --rebuild-daily

Rebuilding the daily summaries

The wee_database utility can also be used to rebuild the daily summaries:

wee_database weewx.conf --rebuild-daily

In most cases this will be sufficient; however, if anomalies remain in the daily summaries the daily summary tables may be dropped first before rebuilding:

wee_database weewx.conf --drop-daily

The summaries will automatically be rebuilt the next time weeWX starts, or they can be rebuilt with the utility:

wee_database weewx.conf --rebuild-daily

Customizing units and unit groups

Warning!
This is an area that is rapidly changing in weeWX. Presently, new units and unit groups are added by manipulating the internal dictionaries in weeWX (as described below). In the future, they may be specified in weewx.conf.

Assigning a unit group

In the examples above, we created a new observation type, electricity, and added it to the database schema. Now we would like to recognize that it is a member of the unit group group_energy (which already exists), so it can enjoy the labels and formats already provided for this group. This is done by extending the dictionary weewx.units.obs_group_dict, typically by adding Python code to the file user/extensions.py:

import weewx.units
weewx.units.obs_group_dict['electricity'] = 'group_energy'

Once the observation has been associated with a unit group, the unit labels and other tag syntax will work for that observation. So, now a tag like:

$month.electricity.sum

will return the total amount of electricity consumed for the month, in Watt-hours.

Creating a new unit group

That was an easy one, because there was already an existing group, group_amp, that covered our new observation type. But, what if we are measuring something entirely new, like force with time? There is nothing in the existing system of units that covers things like newtons or pounds. We will have to define these new units, as well as the unit group they can belong to.

We assume we have a new observation type, force, that we are measuing over time. We will create a new unit group, group_force, and new units, newton and pound. Our new observation, force, will belong to group_force, and will be measured in units of newton or pound.

  1. As before, we start by specifying what group our new observation type belongs to:
    import weewx.units
    weewx.units.obs_group_dict['force'] = 'group_force'
  2. Next, we specify what unit is used to measure force in the three standard unit systems used by weewx.
    weewx.units.USUnits['group_force'] = 'pound'
    weewx.units.MetricUnits['group_force'] = 'newton'
    weewx.units.MetricWXUnits['group_force'] = 'newton'
  3. Then we specify what formats and labels to use for newton and pound:
    weewx.units.default_unit_format_dict['newton'] = '%.1f'
    weewx.units.default_unit_format_dict['pound']  = '%.1f'
    
    weewx.units.default_unit_label_dict['newton'] = ' newton'
    weewx.units.default_unit_label_dict['pound']  = ' pound'
  4. Finally, we specify how to convert between them:
    weewx.units.conversionDict['newton'] = {'pound':  lambda x : x * 0.224809}
    weewx.units.conversionDict['pound']  = {'newton': lambda x : x * 4.44822}
    

Use the new types

Now you've added a new type. How do you use it?

Pretty much like any other type. For example, to do a plot of the month's electric consumption, totaled by day, add this section to the [[month_images]] section of skin.conf:

[[[monthelectric]]]
    [[[[electricity]]]]
        aggregate_type = sum
        aggregate_interval = 86400
        label = Electric consumption (daily total)

This will cause the generation of an image monthelectric.png, showing a plot of each day's consumption for the past month.

If you wish to use the new type in the templates, it will be available using the same syntax as any other type. Here are some other tags that might be useful:

Tag Meaning
$day.electricity.sum Total consumption since midnight
$year.electricity.sum Total consumption since the first of the year
$year.electricity.max The most consumed during any archive period
$year.electricity.maxsum The most consumed during a day
$year.electricity.maxsumtime The day it happened.
$year.electricity.sum_ge(5000.0) The number of days of the year where more than 5.0 kWh of energy was consumed.

Porting to new hardware

Naturally, this is an advanced topic but, nevertheless, I'd like to encourage any Python wizards out there to give it a try. Of course, I have selfish reasons for this: I don't want to have to buy every weather station ever invented, and I don't want my roof to look like a weather station farm!

General guidelines

Implement the driver

Create a file in the user directory, say mydriver.py. This file will contain the driver class as well as any hardware-specific code. Do not put it in the weewx/drivers directory or it will be deleted when you upgrade weeWX.

Inherit from the abstract base class weewx.drivers.AbstractDevice. Try to implement as many of its methods as you can. At the very minimum, you must implement the first three methods, loader, hardware_name, and genLoopPackets.

loader

This is a factory function that returns an instance of your driver. It has two arguments: the configuration dictionary, and a reference to the weeWX engine.

hardware_name

Return a string with a short nickname for the hardware, such as "ACME X90"

genLoopPackets

This should be a generator function that yields loop packets, one after another. Don't worry about stopping it: the engine will do this when an archive record is due. A "loop packet" is a dictionary. At the very minimum it must contain keys for the observation time and for the units used within the packet.

Required keys
dateTime The time of the observation in unix epoch time.
usUnits The unit system used. weewx.US for US customary, weewx.METRICWX, or weewx.METRIC for metric. See the file units.py, dictionaries USUnits, MetricWXUnits, and MetricUnits for the exact definition of each.

Then include any observation types you have in the dictionary. Every packet need not contain the same set of observation types. Different packets can use different unit systems, but all observations within a packet must use the same unit system. If your hardware has an error and you don't have a value, you can either leave it out of the dictionary or (preferred) set its value to None.

A couple of observation types are tricky. In particular, rain. Generally, weeWX expects to see a packet with the amount of rain that fell in that packet period included as observation rain. It then sums up all the values to get the total rainfall and emits that in the archive record. If your hardware does not provide this value, you might have to infer it from changes in whatever value it provides, for example changes in the daily or monthly rainfall. I know this is not the best solution, but it is the most general solution. Any alternatives are welcome!

Wind is another tricky one. It is actually broken up into four different observations: windSpeed, windDir, windGust, and windGustDir. Supply as many as you can. The directions should be compass directions in degrees (0=North, 90=East, etc.).

Be careful when reporting pressure. There are three observations related to pressure. Some stations report only the station pressure, others calculate and report sea level pressures.

Pressure types
pressure The Station Pressure (SP), which is the raw, absolute pressure measured by the station. This is the true barometric pressure for the station.
barometer The Sea Level Pressure (SLP) obtained by correcting the Station Pressure for altitude and local temperature. This is the pressure reading most commonly used by meteorologist to track weather systems at the surface, and this is the pressure that is uploaded to weather services by weeWX. It is the station pressure reduced to mean sea level using local altitude and local temperature.
altimeter The Altimeter Setting (AS) obtained by correcting the Station Pressure for altitude. This is the pressure reading most commonly heard in weather reports. It is not the true barometric pressure of a station, but rather the station pressure reduced to mean sea level using altitude and an assumed temperature average.

genArchiveRecords()

If your hardware does not have an archive record logger, then weeWX can do the record generation for you. It will automatically collect all the types it sees in your loop packets then emit a record with the averages (in some cases the sum or max value) of all those types. If it doesn't see a type, then it won't appear in the emitted record.

However, if your hardware does have a logger, then you should implement method genArchiveRecords() as well. It should be a generator function that returns all the records since a given time.

archive_interval

If you implement function genArchiveRecords() above, then you should also implement archive_interval as either an attribute, or as a property function. It should return the archive interval in seconds.

getTime()

If your hardware has an onboard clock and supports reading the time from it, then you may want to implement this method. It takes no argument. It should return the time in Unix Epoch Time.

setTime()

If your hardware has an onboard clock and supports setting it, then you may want to implement this method. It takes no argument and does not need to return anything.

closePort()

If the driver needs to close a serial port, terminate a thread, close a database, or perform any other activity before the application terminates, then you must supply this function. WeeWX will call it if it needs to shut down your console (usually in the case of an error).

Define the configuration

You then include a new section in the configuration file weewx.conf that includes any options your driver needs. It should also include an entry driver that points to where your driver can be found. Set option station_type to your new section type and your driver will be loaded.

Examples

The fileparse driver is perhaps the most simple example of a weeWX driver. It reads name-value pairs from a file and uses the values as sensor 'readings'. The code is located in extensions/fileparse/bin/user/fileparse.py

Take a look at the simulator code in bin/weewx/drivers/simulator.py. It's dirt simple and you can easily play with it. Many people have successfully used it as a starting point for writing their own custom driver.

The Ultimeter (ultimeter.py) and WMR100 (wmr100.py) drivers illustrate how to communicate with serial and USB hardware, respectively. They also show different approaches for decoding data. Nevertheless, they are pretty straightforward.

The driver for the Vantage series is by far the most complicated. It actually multi-inherits from not only AbstractDevice, but also StdService. That is, it also participates in the engine as a service.

Naturally, there are a lot of subtleties that I've glossed over in this high-level description. If you're game, give it a try — I'm happy to help you out!

Extensions

A key feature of weeWX is its ability to be extended by installing 3rd party extensions. Extensions are a way to package one or more customizations so they can be installed and distributed as a functional group.

Customizations typically fall into one of these categories:

Take a look at the weeWX wiki for a sampling of some of the extensions that are available.

Creating an extension

Now that you have made some customizations, you might want to share those changes with other weeWX users. Put your customizations into an extension to make installation, removal, and distribution easier.

Here are a few guidelines for creating extensions:

How to package an extension

The structure of an extension mirrors that of weeWX itself. If the customizations include a skin, the extension will have a skins directory. If the customizations include python code, the extension will have a bin/user directory.

Each extension should also include:

For example, here is the structure of a skin called basic:

basic/
basic/changelog
basic/install.py
basic/readme.txt
basic/skins/
basic/skins/basic/
basic/skins/basic/basic.css
basic/skins/basic/current.inc
basic/skins/basic/favicon.ico
basic/skins/basic/hilo.inc
basic/skins/basic/index.html.tmpl
basic/skins/basic/skin.conf

Here is the structure of a search list extension called xstats:

xstats/
xstats/changelog
xstats/install.py
xstats/readme.txt
xstats/bin/
xstats/bin/user/
xstats/bin/user/xstats.py

See the extensions directory of the weeWX source for examples.

To distribute an extension, simply create a compressed archive of the extension directory.

For example, create the compressed archive for the basic skin like this:

tar cvfz basic.tar.gz basic

Once an extension has been packaged, it can be installed using the wee_extension utility.

Appendix: Archive Types

Archive types are weather observations that have come from your instrument and been stored in the archive database, a SQL database. They represent the current conditions as of some time. They are available to be used in two places:

The following table shows all the possible archive types and whether they can be used in tag $current or in a plot. Note that just because a type appears in the table does not necessarily mean that it is available for your station setup. That would depend on whether your instrument supports the type.

Archive types
Archive Type SQL Type
(appears in archive database)
Can be used
in plots
Can be used
in tag $current
altimeter X X X
barometer X X X
consBatteryVoltage X X X
dateTime X
X (represents current time)
dewpoint X X X
ET X X X
extraHumid1 X X X
extraHumid2 X X X
extraTemp1 X X X
extraTemp2 X X X
extraTemp3 X X X
hail X X X
hailRate X X X
heatindex X X X
heatingTemp X X X
heatingVoltage X X X
inHumidity X X X
inTemp X X X
inTempBatteryStatus X X X
interval X X X
leafTemp2 X X X
leafWet2 X X X
outHumidity X X X
outTemp X X X
outTempBatteryStatus X X X
pressure X X X
radiation X X X
rain X X X
rainBatteryStatus X X X
rainRate X X X
referenceVoltage X X X
rxCheckPercent X X X
soilMoist1 X X X
soilMoist2 X X X
soilMoist3 X X X
soilMoist4 X X X
soilTemp1 X X X
soilTemp2 X X X
soilTemp3 X X X
soilTemp4 X X X
supplyVoltage X X X
txBatteryStatus X X X
usUnits X X X
UV X X X
windvec   X (special vector type)  
windBatteryStatus X X X
windDir X X X
windGust X X X
windGustDir X X X
windSpeed X X X
windchill X X X

Appendix: Aggregation types

Aggregation types
Aggregation type Meaning
avg The average value in the aggregation period.
sum The sum of values in the aggregation period.
count The number of non-null values in the aggregation period.
min The minimum value in the aggregation period.
mintime The time of the minimum value.
max The maximum value in the aggregation period.
maxtime The time of the maximum value.
maxmin The maximum daily minimum in the aggregation period. Aggregation period must be one day or longer.
maxmintime The time of the maximum daily minimum.
minmax The minimum daily maximum in the aggregation period. Aggregation period must be one day or longer.
minmaxtime The time of the minimum daily maximum.
maxsum The maximum daily sum in the aggregation period. Aggregation period must be one day or longer.
maxsumtime The time of the maximum daily sum.
meanmin The average daily minimum in the aggregation period. Aggregation period must be one day or longer.
meanmax The average daily maximum in the aggregation period. Aggregation period must be one day or longer.
gustdir The direction of the max gust in the aggregation period.
last The last value in the aggregation period.
lasttime The time of the last value in the aggregation period.
max_ge(val) The number of days where the maximum value is greater than or equal to val. Aggregation period must be one day or longer.
max_le(val) The number of days where the maximum value is less than or equal to val. Aggregation period must be one day or longer.
min_ge(val) The number of days where the minimum value is greater than or equal to val. Aggregation period must be one day or longer.
min_le(val) The number of days where the minimum value is less than or equal to val. Aggregation period must be one day or longer.
sum_ge(val) The number of days where the sum of value is greater than or equal to val. Aggregation period must be one day or longer.
rms The root mean square value in the aggregation period.
vecavg The vector average speed in the aggregation period.
vecdir The vector averaged direction during the aggregation period.

Appendix: Units

WeeWX offers three different unit systems:

The standard unit systems used within weeWX
Name Encoded value Note
US 0x01 U.S. Customary
METRICWX 0x11 Metric, with rain related measurements in mm and speeds in m/s
METRIC 0x10 Metric, with rain related measurements in cm and speeds in km/hr

The table below lists all the unit groups, their members, which units are options for the group, and what the defaults are for each standard unit system.

Unit groups, members and options
Group Members Unit options US METRICWX METRIC
group_altitude altitude
cloudbase
foot
meter
foot meter meter
group_amp amp amp amp amp
group_data byte
bit
byte byte byte
group_degree_day cooldeg
heatdeg
degree_F_day
degree_C_day
degree_F_day degree_C_day degree_C_day
group_direction gustdir
vecdir
windDir
windGustDir
degree_compass degree_compass degree_compass degree_compass
group_distance windrun mile
km
mile km km
group_energy watt_hour watt_hour watt_hour watt_hour
group_interval interval minute minute minute minute
group_length inch
cm
inch cm cm
group_moisture soilMoist1
soilMoist2
soilMoist3
soilMoist4
centibar centibar centibar centibar
group_percent extraHumid1
extraHumid2
inHumidity
outHumidity
rxCheckPercent
percent percent percent percent
group_power watt watt watt watt
group_pressure barometer
altimeter
pressure
inHg
mbar
hPa
inHg mbar mbar
group_radiation radiation watt_per_meter_squared watt_per_meter_squared watt_per_meter_squared watt_per_meter_squared
group_rain rain
ET
hail
inch
cm
mm
inch mm cm
group_rainrate rainRate
hailRate
inch_per_hour
cm_per_hour
mm_per_hour
inch_per_hour mm_per_hour cm_per_hour
group_speed wind
windGust
windSpeed
windgustvec
windvec
mile_per_hour
km_per_hour
knot
meter_per_second
mile_per_hour meter_per_second km_per_hour
group_speed2 rms
vecavg
mile_per_hour2
km_per_hour2
knot2
meter_per_second2
mile_per_hour2 meter_per_second2 km_per_hour2
group_temperature appTemp
dewpoint
extraTemp1
extraTemp2
extraTemp3
heatindex
heatingTemp
humidex
inTemp
leafTemp1
leafTemp2
outTemp
soilTemp1
soilTemp2
soilTemp3
soilTemp4
windchill
degree_F
degree_C
degree_F degree_C degree_C
group_time dateTime unix_epoch
dublin_jd
unix_epoch unix_epoch unix_epoch
group_uv UV uv_index uv_index uv_index uv_index
group_volt consBatteryVoltage
heatingVoltage
referenceVoltage
supplyVoltage
volt volt volt volt
group_volume cubic_foot
gallon
litre
gallon litre litre
group_NONE NONE NONE NONE NONE NONE