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 is 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

Where are waterbodies located in Vienna?

Again, we have to resolve the name ‘Vienna’ before running the query:

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

The query can be built like in the examples before. This time, however, the argument includeGeometry=True is provided to the overpassQueryBuilder in order to let him generate a query that downloads the geometry data.

from OSMPythonTools.overpass import overpassQueryBuilder, Overpass
overpass = Overpass()
query = overpassQueryBuilder(area=areaId, elementType=['way', 'relation'], selector='"natural"="water"', includeGeometry=True)
result = overpass.query(query)

Next, we can exemplarily choose one random waterbody (the first one of the download ones) and compute its geometry like follows:

firstElement = result.elements()[0]
firstElement.geometry()
# {"coordinates": [[[16.498671, 48.27628], [16.4991, 48.276345], ... ]], "type": "Polygon"}

Observe that the resulting geometry is provided in the GeoJSON format.

Example 5

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')

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.

Usage

The following modules are available (please click on their names to access further documentation):

• OSMPythonTools.Api - Access to the official OSM API
• OSMPythonTools.Data - Collecting, mining, and drawing data from OSM; to be used in combination with the other modules
• OSMPythonTools.Element - Elements are returned by other services, like the OSM API or the Overpass API
• OSMPythonTools.Nominatim - Access to Nominatim, a reverse geocoder
• OSMPythonTools.Overpass - Access to the Overpass API

Logging

This library is a little bit more verbose than other Python libraries. The good reason behind is that the OpenStreetMap, the Nominatim, and the Overpass servers experience a heavy load already and their resources should be used carefully. In order to make you, the user of this library, aware of when OSMPythonTools accesses these servers, corresponding information is logged by default. In case you want to suppress these messages, you have to insert the following lines after the import of OSMPythonTools:

import logging
logging.getLogger('OSMPythonTools').setLevel(logging.ERROR)

Please note that suppressing the messages means that you have to ensure on your own that you do not overuse the provided services and that you stick to their fair policy guidelines.

Tests

You can test the package by running

pytest --verbose

Please note that the tests might run very long (several minutes) because the overpass server will most likely defer the downloads.

Author

This application is written and maintained by Franz-Benjamin Mocnik, [email protected].

(c) by Franz-Benjamin Mocnik, 2017-2020.

The code is licensed under the GPL-3.