tracker is sofware for analyzing and displaying Garmin .tcx files.

If you have a device that uses the Garmin ANT stick, you can obtain the data by
running gant from the subdirectory gant. Look at the README in there for more
information. The gant downloading is a little flaky. If your watch fails to
download the latest tracks, you'll have to select Force->Send, which will then
send all the tracks stored on the device, which will take quite some time. Also,
the downloads can be a bit temperamental, so be patient. It is best to have the
watch charging when you are downloading - that seems to work more reliably. Note
that some of the tracks downloaded will be summarized, as indicated in the file
name. These are older tracks for which the watch has dropped the
trackpoints. Don't replace your previous unsummarized files with summarized
ones: you will lose all your trackpoints and there is no way of getting them
back.

Once you have all the .tcx files, you can run tracker.py to analyze them.

For options, run:

tracker.py -h 

Usually, the first thing to do is to update the tracker database from the xml
files, i.e.

tracker.py data/*.tcx

The data will be stored in a file (tracks.pck by default) and that will be used
in future. It is just used to speed things up, you can always delete it and
simply rescan the .tcx files.

Within the .tcx files, tracker can read a few additional fields that don't exist
in the Garmin spec. These should be placed at the same level as the <Id> field,
i.e. just below <Activity>. The fields are:

1. add comments with <Comment>text</Comment>. The text can be anything. New
lines are respected and all comments will be printed out as formatted when
printing tracks.
   
2. add time zone with <TimeZone>time_zone</TimeZone>. There is also the option
to specify the timezone on the command line, with the -z switch. If you pass it
-z auto, then tracker tries to infer the timezone from the lat and lng of the
first trackpoint. If there are no trackpoints, then it just uses UTC,
i.e. GMT. The timezone inference is crude and won't take into account such
things as daylight savings or accurate country/state boundaries. More work is
needed on the auto option. In general, the inference is seldom more than an hour
out. 

3. add known distance and elevation with
<KnownElevationMeters>known_elev</KnownElevationMeters>, and
<KnownDistanceMeters>known_dist</KnownDistanceMeters>. Use these if you know the
actual distance or elevation for the track. This is particularly useful for
summarized runs for which you don't have the trackpoints to calculate the
elevation. These values are only used if there are no GPS distances or
elevations. 

In addition to these fields, tracker will also add a field <MapAltitudeMeters>
to every trackpoint on the first scan and write it into the .tcx file. tracker
obtains this information by submitting a request to Google for elevations
corresponding to lat,lng coords and is more accurate than using the GPS
elevations.

For example, to load a new track (say 2011-06-20-174608) in timezone US/Pacific,
and map it on Google, run

tracker.py -m 2011-06-20-174608 -z US/Pacific data/2011-06-20-174608.tcx

For most analysis, tracker will both plot data and list data as text
output. For plots of daily information, it doesn't yet handle multiple runs on
one day properly. 

Please feel free to play around and let me know of any fixes/feedback.