tcxreader is a reader for Garminโ€™s TCX file format. It also works well with missing data!


Keywords
data-mining, data-science, python, sports-analytics, tcx, tcx-parser
License
MIT
Install
pip install tcxreader==0.4.1

Documentation

tcxreader is a reader for Garmin's TCX file format. It also works well with missing data!


PyPI Version PyPI - Python Version PyPI - Downloads Downloads GitHub repo size GitHub license GitHub commit activity Average time to resolve an issue Percentage of issues still open All Contributors

Objective

This is a simple TCX reader which can read Garmin TCX file extension files. The following data is currently parsed: longitude, latitude, elevation, time, distance, hr_value, cadence, watts, TPX_speed (extension). The following statistics are calculated for each exercise: calories, hr_avg, hr_max, hr_min, avg_speed, start_time, end_time, duration, cadence_avg, cadence_max, ascent, descent, distance, altitude_max, altitude_min, altitude_avg, steps and author data.

GitHub requests appreciated. pypi github

Features

Allows parsing / reading of TCX files.

Installation

pip install tcxreader

Example

An example on how to use the package is shown below.

from tcxreader.tcxreader import TCXReader, TCXTrackPoint

tcx_reader = TCXReader()
file_location = 'example_data/cross-country-skiing_activity_1.tcx'

"""
Minor warning, the read method also has a default parameter of only_gps (tcx_readerread(self, fileLocation: str, only_gps: bool = True)) set to true. If set to True erases any Trackpoints at the start and end of the exercise without GPS data.
"""

data: TCXExercise = tcx_reader.read(file_location)
""" Example output:
data = {TCXExercise}
 activity_type = {str} 'Other'
altitude_avg = {float} 2285.6744874553915
altitude_max = {float} 2337.60009765625
altitude_min = {float} 1257.5999755859375
ascent = {float} 1117.9996337890625
author = {TCXAuthor} [TCXAuthor]
avg_speed = {float} 8.534458975426906
cadence_avg = {NoneType} None
cadence_max = {NoneType} None
calories = {int} 532
descent = {float} 118.19970703125
distance = {float} 5692.01
duration = {float} 2401.0
end_time = {datetime} 2020-12-26 15:54:22
hr_avg = {float} 141.1954732510288
hr_max = {int} 172
hr_min = {int} 83
laps = {list: 2} [TCXLap]
lx_ext = {dict: 0} {}
max_speed = {float} 23.50810546875
start_time = {datetime} 2020-12-26 15:14:21
tpx_ext_stats = {dict: 2} {'Speed': {'min': 0.0, 'max': 6.1579999923706055, 'avg': 2.2930514418784482}, 'RunCadence': {'min': 0, 'max': 95, 'avg': 40.81069958847737}}
trackpoints = {list: 486} [TCXTrackpoint]
 
    {TCXTrackPoint} 
     cadence = {NoneType} None
     distance = {float} 7.329999923706055
     elevation = {float} 2250.60009765625
     hr_value = {int} 87
     latitude = {float} 46.49582446552813
     longitude = {float} 15.50408081151545
     time = {datetime} 2020-12-26 15:14:28
     tpx_ext = {dict: 2} {'Speed': 0.7459999918937683, 'RunCadence': 58}
"""

Classes explanation

Below figure explains the classes of tcxreader and the data they contain.

TCXReader()

User initializes the tcxreader by creating a TCXReader class instance. To read the data of a TCX activity the user must use TCXReader.read(filename) method. The output of read() is an instance of TCXExercise class.

TCXExercise

Primary class that holds cumulative data of an exercise. TCXExercise contains all the trackpoints of an activity (e.g. from all the laps merged).

TCXLap

One TCX activity may contain multiple laps. In the TCX file they are visible by the Lap tag.

<Lap StartTime="2020-12-26T15:50:22.000Z">
...
</Lap>

TCXLap contains all the trackpoints of a lap.

TCXTrackpoint

A point in an exercise. Almost always has latitude, longitude, time. Can also have cadence, distance, elevation, hr_value, tpx_ext. The tpx_ext refers to individual extensions contained inside the trackpoint. An example of the Trackpoint (pre-parsing) in the TCX file is shown below.

<Trackpoint>
    <Time>2020-12-26T15:50:21.000Z</Time>
    <Position>
        <LatitudeDegrees>46.49732105433941</LatitudeDegrees>
        <LongitudeDegrees>15.496849408373237</LongitudeDegrees>
    </Position>
    <AltitudeMeters>2277.39990234375</AltitudeMeters>
    <DistanceMeters>5001.52978515625</DistanceMeters>
    <HeartRateBpm>
        <Value>148</Value>
    </HeartRateBpm>
    <Extensions>
        <ns3:TPX>
            <ns3:Speed>3.3589999675750732</ns3:Speed>
            <ns3:RunCadence>61</ns3:RunCadence>
        </ns3:TPX>
    </Extensions>
</Trackpoint>

tpx_ext

The data parsed from the trackpoint TPX Extensions. Example of data (pre-parsing) is shown below.

<Extensions>
    <ns3:TPX>
        <ns3:Speed>3.3589999675750732</ns3:Speed>
        <ns3:RunCadence>61</ns3:RunCadence>
    </ns3:TPX>
</Extensions>

Can occur once (1x) in every trackpoint.

tpx_ext_stats

Contains minimum, maximum and average values of the recorded tpx_ext key.

lx_ext

The data parsed from the lap LX Extensions. Example of data (pre-parsing) is shown below.

<Extensions>
    <ns3:LX>
        <ns3:AvgSpeed>1.0820000171661377</ns3:AvgSpeed>
        <ns3:Steps>65</ns3:Steps>
    </ns3:LX>
</Extensions>

Can occur once (1x) in every lap.

The tags which do not contain Avg, Min, Max in their name (e.g. steps) are summed in the TCXExercise lx_ext dictionary.

All tags are recorded in the TCXLap lx_ext dictionary

Schema of the data

Missing data handling

Due to the nature of the TCX file format, some data may be missing. The tcxreader can handle this in two ways:

  1. If data is missing at a TCX point it is set to None. (default)
    • tcx_reader.read(file_location) (default)
    • tcx_reader.read(file_location, null_value_handling=1) (default)
    • tcx_reader.read(file_location, null_value_handling=NullValueHandling.NONE) (default)
  2. If data is missing at one or more TCX points it is linearly interpolated.
    • tcx_reader.read(file_location, null_value_handling=2)
    • tcx_reader.read(file_location, null_value_handling=NullValueHandling.LINEAR_INTERPOLATION)

This behavior can be set in TCXReader.read() method by the null_value_handling parameter, where either int value or NullValueHandling enum can be passed.

Datasets

Datasets available and used in the examples on the following links: DATASET1, DATASET2, DATASET3.

License

This package is distributed under the MIT License. This license can be found online at http://www.opensource.org/licenses/MIT.

Related packages/frameworks

[1] sport-activities-features: A minimalistic toolbox for extracting features from sports activity files written in Python

[2] AST-Monitor: A wearable Raspberry Pi computer for cyclists

Disclaimer

This framework is provided as-is, and there are no guarantees that it fits your purposes or that it is bug-free. Use it at your own risk!

Contributors

alenrajsp
alenrajsp

๐Ÿ’ป ๐Ÿšง
fortysix2ahead
fortysix2ahead

๐Ÿ›
Iztok Fister Jr.
Iztok Fister Jr.

๐Ÿ”ฃ ๐Ÿง‘โ€๐Ÿซ ๐Ÿ“ฆ โš ๏ธ
johnleeming
johnleeming

๐Ÿ›
rpstar
rpstar

๐Ÿ›
James Robinson
James Robinson

๐Ÿšง
johwiebe
johwiebe

๐Ÿ›
Martin Ueding
Martin Ueding

๐Ÿ›
Simon Pickering
Simon Pickering

๐Ÿ›
Rich Winkler
Rich Winkler

๐Ÿ›