The TrackAddict Command Line Interface (CLI) will help to slice and dice the raw csv output from the TrackAddict app. Currently it has functionality to smooth GPS data, recalculate laps and plot them on an OpenStreet map.
We built this tool to fix bad GPS resolution/reception from Android phones and to still create some reliable lap times out of noisy measurements.
Disclaimer: This project is not affiliated with the TrackAddict product.
With golang 1.12 (or higher) installed, just run
go install github.com/thomasjungblut/trackaddict-cli
Make sure your PATH includes the $GOPATH/bin directory so your commands can be easily used:
export PATH=$PATH:$GOPATH/bin
Printing the Laps that TrackAddict has created can be done via:
trackaddict-cli laps -i example/STC_log.csv
That should give an output like this:
+------------+-----------+---------------+
| LAP NUMBER | TIME (S) | MEASURE RANGE |
+------------+-----------+---------------+
| 1 (Outlap) | 2m35.214s | 0-3154 |
| 2 | 7m23.887s | 3155-12133 |
| 3 | 1m47.963s | 12134-14317 |
| 4 | 1m55.877s | 14318-16665 |
| 5 | 5m57.949s | 16666-23896 |
| 6 (Inlap) | 1m4.231s | 23897-25196 |
+------------+-----------+---------------+
As you can see here, some laps seem to get mixed together by noisy GPS measures, let's plot them to visualize:
trackaddict-cli plot -i example/STC_log.csv -o docs/raw_output.png
That's pretty noisy! Let's try to smooth the GPS data a bit. We're using the accelerometer data to improve the GPS signal by using a Kalman Filter. You can have all of that power with just one single flag:
trackaddict-cli plot -i example/STC_log.csv -o docs/filtered_output.png --smooth
As you can see, especially the bits where the accelerometer has strong evidence (eg after the long straight and before the hairpin turn on the top left), the GPS signal became much more accurate.
The same works for the lap times. Let's smooth the GPS data again and recalculate the laps based on that:
trackaddict-cli laps -i example/STC_log.csv --smooth --fix-laps
We can also plot laps individually:
trackaddict-cli plot -i example/STC_log.csv -o docs/lap_plot --smooth --fix-laps --plot-each-lap
smooth and fix flags are optional, so you can also plot the track addict data directly if your data does not need any filtering:
trackaddict-cli plot -i example/STC_log.csv -o docs/lap_plot --plot-each-lap
That gives us the ability to just take a look at our inlap at the end for example:
