OSMPythonTools

From OpenStreetMap Wiki
Jump to: navigation, search
OSMPythonTools
Screenshot of OSMPythonTools
Author: Franz-Benjamin Mocnik
License: GNU General Public License v3 (free of charge)
Platforms: Python, Windows, macOS, and Linux
Version: 0.2.2 (2018-07-20)
Language: English
Website: https://github.com/mocnik-science/osm-python-tools
Source code: https://github.com/mocnik-science/osm-python-tools
Programming language: Python

Python 3 API for Overpass, Nominatium, and the OSM API

OSMPythonTools

The python package OSMPythonTools provides easy access to OpenStreetMap (OSM) related services, among them an Overpass endpoint, Nominatim, and the OSM API.

Installation

To install OSMPythonTools, you will need python3 and pip (How to install pip). Then execute:

pip install OSMPythonTools

On some operating systems, pip for python3 will be named pip3:

pip3 install OSMPythonTools

Example 1

Which object does the way with the id 5887599 represent?

We can use the OSM API to answer this question:

from OSMPythonTools.api import Api
api = Api()
way = api.query('way/5887599')

The resulting object contains information about the way, which can easily be accessed:

way.tag('building')
# 'castle'
way.tag('architect')
# 'Johann Lucas von Hildebrandt'
way.tag('website')
# 'http://www.belvedere.at'

Example 2

What is the English name of the church called "Stephansdom", what address does it have, and which of which denomination is the church?

We use the Overpass API to query the corresponding data:

from OSMPythonTools.overpass import Overpass
overpass = Overpass()
result = overpass.query('way["name"="Stephansdom"]; out body;')

This time, the result is a number of objects, which can be accessed by result.elements(). We just pick the first one:

stephansdom = result.elements()[0]

Information about the church can now easily be accessed:

stephansdom.tag('name:en')
# "Saint Stephen's Cathedral"
'%s %s, %s %s' % (stephansdom.tag('addr:street'), stephansdom.tag('addr:housenumber'), stephansdom.tag('addr:postcode'), stephansdom.tag('addr:city'))
# 'Stephansplatz 3, 1010 Wien'
stephansdom.tag('building')
# 'cathedral'
stephansdom.tag('denomination')
# 'catholic'

Example 3

How many trees are in the OSM data of Vienna? And how many trees have there been in 2013?

This time, we have to first resolve the name "Vienna" to an area id:

from OSMPythonTools.nominatim import Nominatim
nominatim = Nominatim()
areaId = nominatim.query('Vienna, Austria').areaId()

This area id can now be used to build the corresponding query:

from OSMPythonTools.overpass import overpassQueryBuilder, Overpass
overpass = Overpass()
query = overpassQueryBuilder(area=areaId, elementType='node', selector='"natural"="tree"', out='count')
result = overpass.query(query)
result.countElements()
# 137830

There are 134520 trees in the current OSM data of Vienna. How many have there been in 2013?

result = overpass.query(query, date='2013-01-01T00:00:00Z', timeout=60)
result.countElements()
# 127689

Example 4

How did the number of trees in Berlin, Paris, and Vienna change over time?

Before we can answer the question, we have to import some modules:

from collections import OrderedDict
from OSMPythonTools.data import Data, dictRangeYears, ALL
from OSMPythonTools.overpass import overpassQueryBuilder, Overpass

The question has two "dimensions": the dimension of time, and the dimension of different cities:

dimensions = OrderedDict([
    ('year', dictRangeYears(2013, 2017.5, 1)),
    ('city', OrderedDict({
        'berlin': 'Berlin, Germany',
        'paris': 'Paris, France',
        'vienna': 'Vienna, Austria',
    })),
])

We have to define how we fetch the data. We again use Nominatim and the Overpass API to query the data (it can take some time to perform this query the first time!):

overpass = Overpass()
def fetch(year, city):
    areaId = nominatim.query(city).areaId()
    query = overpassQueryBuilder(area=areaId, elementType='node', selector='"natural"="tree"', out='count')
    return overpass.query(query, date=year, timeout=60).countElements()
data = Data(fetch, dimensions)

We can now easily generate a plot from the result:

data.plot(city=ALL, filename='example4.png')
File:OSMPythonTools/example4.png
caption data.plot(city=ALL, filename='example4.png')

Alternatively, we can generate a table from the result

data.select(city=ALL).getCSV()
# year,berlin,paris,vienna
# 2013.0,10180,1936,127689
# 2014.0,17971,26905,128905
# 2015.0,28277,90599,130278
# 2016.0,86769,103172,132293
# 2017.0,108432,103246,134616

More examples can be found inside the documentation of the modules.

Documentation

For further documentation, please refer to (https://github.com/mocnik-science/osm-python-tools)[1].