About

Load GPS data from GPX files into Python as a numpy arrays and pandas DataFrames. Initial parsing done using the gpxpy package. Trajectory plotting on a map available using mplleaflet (optional).

Quick Start

Install

pip install gpxo

Load Track

import gpxo
track = gpxo.Track('ExampleTrack.gpx')

(it is possible to indicate which track or segment to consider during instantiation, by default it is the first one). track.data is a pandas DataFrame containing time, position, elevation etc.; usual pandas methods can be used to analyze, manipulate and plot data. Individual columns are also available as numpy arrays as attributes of the class (see below).

Detailed Contents

Track class

Load, inspect and plot GPX data using the Track class, with the following methods and attributes.

Methods

• smooth(): smooth position and elevation data (see gpxo.smooth() below),
• plot(): plot trajectory data using a combination of shortnames (see shortnames below); also takes matplotlib.pyplot.plot() arguments/kwargs,
• map(): plot trajectory on a map, using mplleaflet.show(),
• closest_to(): find index of point in trajectory closest to a (lat, long) point.

Basic Attributes

(some may not be available depending on actual data present in the GPX file)

• latitude (numpy array): latitude in °,
• longitude (numpy array): longitude in °,
• elevation (numpy array): elevation in meters,
• time (numpy array): local time expressed as a datetime.datetime.

Property attributes

(Read-only, and calculated/updated from basic attributes; some may not be available depending on actual data present in the GPX file)

• seconds (numpy array): total number of seconds since beginning of track,
• distance (numpy array): total distance (km) since beginning of track,
• velocity (numpy array): instantaneous velocity (km/h),
• compass (numpy array): instantaneous compass bearing (°),
• data (pandas DataFrame): all above attributes in a single dataframe.

Miscellaneous

Outside of the Track class, the following standalone function is also available:

• compass(pt1, pt2): compass bearing (°) between pt1 (lat1, long1) and pt2 (lat2, long2),
• closest_pt(pt, trajectory): index of closest pt in trajectory (latitudes, longitudes) to specified pt (lat, long),
• smooth(x, n, window): smooth 1-d array with a moving window of size n and type window.

Short names

Short name	Corresponding data
t	time
s	duration (s)
d	distance (km)
v	velocity (km/h)
z	elevation (m)
c	compass (°)

Examples

See Jupyter Notebook Examples.ipynb (https://github.com/ovinc/gpxo/blob/master/Examples.ipynb) for a detailed example using real GPX data.

Quick example 1: Plot distance versus time and velocity versus position for a given track:

import gpxo
track = gpxo.Track('ExampleTrack.gpx')
track.plot('td', '--k')    # matplotlib styles can be given
track.plot('dv', c='red')  # matplotlib kwargs can be passed
track.data  # pandas dataframe with all data

Quick example 2: show the path of a GPX file on a map with color-coding corresponding to elevation (see Troubleshooting section below in case of error):

import gpxo
track = gpxo.Track('ExampleTrack.gpx')
track.map(plot='scatter', c=track.elevation, cmap='plasma')

Troubleshooting

In case of the following error:

'XAxis' object has no attribute '_gridOnMajor

when using the map() method, try downgrading Matplotlib to version <= 3.3.2 or install a forked version of mplleaflet (see jwass/mplleaflet#75).

Information

Requirements

Python >= 3.6

Dependencies

(automatically installed by pip if necessary)

• numpy
• pandas
• matplotlib
• importlib-metadata
• gpxpy (https://github.com/tkrajina/gpxpy)
• vincenty (https://github.com/maurycyp/vincenty)

Optional:

• mplleaflet (https://github.com/jwass/mplleaflet)

Author

Olivier Vincent

(ovinc.py@gmail.com)

License

BSD 3-Clause (see LICENCE file)