rasta

Rasta (rāstā) is a Python based library specifically design to handle geospatial data especially in the context of navigation. Currently, one can use rasta to parse/process/visualize GPX, GTFS, GEOJSON data. In addition, there are modules to use REST API from HERE and OpenTripPlanner for the navigation.

Features

• Parse GPX file and convert them to pandas dataframe in a single line of code. Finally able to visualize the tracks with kepler.gl based interactive map.
• Parse GTFS feed and visualize the network.
• Navigate with HERE Maps or OpenTripPlanned by using their respective REST APIs.

Install

Directly install from PyPI:

pip install rasta

Alternatively, clone this repo and install:

git clone https://github.com/ikespand/rasta.git
cd rasta
python setup.py install

To visualize data with the kepler.gl, user needs to have Mapbox API key which provides the tiles. Go to Mapbox.com to get an API key.

Usage

There are many examples available in the example folder. Additionally, code is documented with docstrings.

Using GPX tracks

Examples gpx_bicycle_tracks.py and gpx_dmrc_yellow_line.py show the usage of gpx module for parsing gpx file along with the data processing. For any generic GPX file:

from rasta.gpx import GpxParser
# Load the gpx file (if timestamp is available then also calculate speed)
gpx_instance = GpxParser("../tracks/BicyleRoute.gpx", calculate_distance=True)
# Extract our data in a dataframe
df = gpx_instance.data
# Visaulize the tracks (You will need Mapbox API key for this step)
html_path, vis = gpx_instance.visualize_route(MAPBOX_API_KEY=MAPBOX_API_KEY,open_browser=True)

Using GTFS feed

Example gtfs_sample_london.py is to have the visalization of GTFS feed from a given zip file. For a generic GTFS feed:

from rasta.gtfs import Gtfs
# Parse our GTFS feed
my_gtfs = Gtfs("../tracks/gtfs_london.zip")
# Visaulize the routes (You will need Mapbox API key for this step)
my_gtfs.visualize_route(MAPBOX_API_KEY, "london")

Interaction with REST API of OpenTripPlanner

otp_visualize_from_rest.ipynb is a notebook to demonstrate the usage of OpenTripPlanner's REST API with rasta. Let's say OTP server is ruuning then one could use:

from navigate_with_otp import GetOtpRoute
import pandas as pd
import numpy as np
# Pass arguments for our query along with parameters for visulization
my_otp_nav = GetOtpRoute(start_coord="28.658420, 77.230757", end_coord="28.544442, 77.206334", MAPBOX_API_KEY=MAPBOX_API_KEY, output_map_path="temporary_map", viz=False)
my_otp_nav.address
gdf, html_path = my_otp_nav.extract_itinerary()

Troubleshooting and some hints

• With keplergl>0.2.1 auto-centering for map data is not working (i.e. you will see San Frasnico as default). Therefore, if you face this issue then consider to downgrade to v0.2.0.
• Sometime, the issue of geopandas, fiona and GDAL appears due to some inter-dependencies. To mitigate it, simply uninstall all 3 and use conda forge to install geopandas.
• If you are new to OTP, then try using docker solution which is documented here.

TODO

• Support for OpenStreetMap data coming from overpy
• GTFS validator
• More tests
• Extend documentation

Developed at: