In [1]:

import os
import folium

print(folium.__version__)

0.6.0+4.gcecbc85.dirty

How to draw a GeoPandas.GeoDataFrame into folium¶

GeoPandas is a project to add support for geographic data to pandas objects. (See https://github.com/geopandas/geopandas)

It provides (among other cool things) a GeoDataFrame object that represents a Feature collection. When you have one, you may be willing to use it on a folium map. Here's the simplest way to do so.

In this example, we'll use the same file as GeoPandas demo ; it's containing the boroughs of New York City.

In [2]:

import geopandas

nybb = os.path.join('data', 'nybb.shp')

boros = geopandas.read_file(nybb)

boros

Out[2]:

	BoroCode	BoroName	Shape_Leng	Shape_Area	geometry
0	5	Staten Island	330454.175933	1.623847e+09	(POLYGON ((970217.0223999023 145643.3322143555...
1	3	Brooklyn	741227.337073	1.937810e+09	(POLYGON ((1021176.479003906 151374.7969970703...
2	4	Queens	896875.396449	3.045079e+09	(POLYGON ((1029606.076599121 156073.8142089844...
3	1	Manhattan	358400.912836	6.364308e+08	(POLYGON ((981219.0557861328 188655.3157958984...
4	2	Bronx	464475.145651	1.186822e+09	(POLYGON ((1012821.805786133 229228.2645874023...

To create a map with these features, simply put them in a GeoJson:

In [3]:

m = folium.Map([40.7, -74], zoom_start=10, tiles='cartodbpositron')

folium.GeoJson(boros).add_to(m)

m.save(os.path.join('results', 'geopandas_0.html'))

m

Out[3]:

Quite easy.

Well, you can also take advantage of your GeoDataFrame structure to set the style of the data. For this, just create a column style containing each feature's style in a dictionnary.

In [4]:

boros['style'] = [
    {'fillColor': '#ff0000', 'weight': 2, 'color': 'black'},
    {'fillColor': '#00ff00', 'weight': 2, 'color': 'black'},
    {'fillColor': '#0000ff', 'weight': 2, 'color': 'black'},
    {'fillColor': '#ffff00', 'weight': 2, 'color': 'black'},
    {'fillColor': '#00ffff', 'weight': 2, 'color': 'black'},
]

boros

Out[4]:

	BoroCode	BoroName	Shape_Leng	Shape_Area	geometry	style
0	5	Staten Island	330454.175933	1.623847e+09	(POLYGON ((970217.0223999023 145643.3322143555...	{'fillColor': '#ff0000', 'weight': 2, 'color':...
1	3	Brooklyn	741227.337073	1.937810e+09	(POLYGON ((1021176.479003906 151374.7969970703...	{'fillColor': '#00ff00', 'weight': 2, 'color':...
2	4	Queens	896875.396449	3.045079e+09	(POLYGON ((1029606.076599121 156073.8142089844...	{'fillColor': '#0000ff', 'weight': 2, 'color':...
3	1	Manhattan	358400.912836	6.364308e+08	(POLYGON ((981219.0557861328 188655.3157958984...	{'fillColor': '#ffff00', 'weight': 2, 'color':...
4	2	Bronx	464475.145651	1.186822e+09	(POLYGON ((1012821.805786133 229228.2645874023...	{'fillColor': '#00ffff', 'weight': 2, 'color':...

In [5]:

m = folium.Map([40.7, -74], zoom_start=10, tiles='cartodbpositron')

folium.GeoJson(boros).add_to(m)

m.save(os.path.join('results', 'geopandas_1.html'))

m

Out[5]:

Folium should work with any object that implements the __geo_interface__ but be aware that sometimes you may need to convert your data to epsg='4326' before sending it to folium.

In [6]:

import fiona
import shapely


fname = os.path.join('data', '2014_08_05_farol.gpx')

with fiona.open(fname, 'r', layer='tracks') as records:
    tracks = [
        shapely.geometry.shape(record['geometry']) for record in records
    ]

track = tracks[0]

m = folium.Map(tiles='cartodbpositron')
folium.GeoJson(track).add_to(m)

m.fit_bounds(m.get_bounds())

m.save(os.path.join('results', 'geopandas_2.html'))

m

Out[6]: