Movementtypes

extract GPS traces from file and analyze them to reveal movement types

Install

movementtypes is available on pypi, https://pypi.python.org/pypi/movementtypes/0.1.1

$ pip install movementtypes

Documentation

http://movementtypes.readthedocs.io/en/latest/

Usage

CSV file format requirements

• The file must be include column headings.
• Character encodings such as BIG-5 have been converted to UTF-8.

Required columns in the CSV file

id, lat, lng, alt, time

Columns in the CSV file

• id: specific id number for identity field
• lat,lng: if the position information is DMS format(Degrees, minutes and seconds), you need to convert to decimal degrees
• time: this value can be YYYY-MM-DD hh:mm:ss or ISO 8601 date and time format

import movementtypes as mvt

df = mvt.mvtypes("data/gps-trajectory.csv")
# using the KDE module to choose an optimal bandwidth
df.optbwKDE()

# check the result
print(df.result)

# using calVelocity
print(df.calVelocity(df.data.iloc[17], df.data.iloc[18][:-1], df.data.iloc[19]))

# classify speed based on the result
df.classifySpeed()
df.export_csv("output.csv")

Public GPS traces on OSM https://www.openstreetmap.org/traces

(gpx2csv can convert the GPX file to valid CSV files)

# path can be your file path or public gpx link
mvt.convert.gpx2csv("https://www.openstreetmap.org/trace/2550408/data", 'osmdata')

df = mvt.mvtypes('osmdata.csv')

df.optbwKDE()
print(df.result)

Command line usage

$ mvtypes --help
Usage: mvtypes [OPTIONS] PATH OUTPUT

  path: file path

  output: output file

Options:
  --threshold INTEGER  the default threshold setting is 15 minutes by
                       clustering the dataset before calculating velocity
  --inepsg INTEGER     input projection, default is 4326
  --outepsg INTEGER    coverting coordinates from inEPSG to outEPSG for
                       calculating the point of velocity, default is 3857
  --help               Show this message and exit.