Python Geovisualization

Interactive Jupyter Notebook

This notebook provides an introduction to Geovisualization using Python. Both the Matplotlib family and HTML-based visualization are covered. This notebook was created by Becky Vandewalle based off of prior work by Dandong Yin.


Visualization is an important technique to be familiar with to visualize geospatial data and analysis results. There are quite a few ways to visualize geospatial data using Python, both using extensions to existing plotting capabilities and through specialized geospatial libraries.

Some useful documentation is listed here:

Official introductory tutorials for Matplotlib
A relatively-short Matplotlib tutorial
Quick Matplotlibgraphic reference
All basemap methods:
Basemap examples:
Cartopy library docs (advanced plotting based on Matplotlib and basemap)


Run this cell for the rest of the notebook to work!

In [1]:
!pip install --user mapclassify
# aftrer installation, you need to restart kernel
Requirement already satisfied: mapclassify in /opt/conda/lib/python3.7/site-packages (2.1.1)
Requirement already satisfied: pandas in /opt/conda/lib/python3.7/site-packages (from mapclassify) (0.25.3)
Requirement already satisfied: scipy>=0.11 in /opt/conda/lib/python3.7/site-packages (from mapclassify) (1.3.1)
Requirement already satisfied: scikit-learn in /opt/conda/lib/python3.7/site-packages (from mapclassify) (0.21.3)
Requirement already satisfied: deprecated in /opt/conda/lib/python3.7/site-packages (from mapclassify) (1.2.7)
Requirement already satisfied: numpy>=1.3 in /opt/conda/lib/python3.7/site-packages (from mapclassify) (1.17.3)
Requirement already satisfied: pytz>=2017.2 in /opt/conda/lib/python3.7/site-packages (from pandas->mapclassify) (2019.3)
Requirement already satisfied: python-dateutil>=2.6.1 in /opt/conda/lib/python3.7/site-packages (from pandas->mapclassify) (2.8.1)
Requirement already satisfied: joblib>=0.11 in /opt/conda/lib/python3.7/site-packages (from scikit-learn->mapclassify) (0.14.0)
Requirement already satisfied: wrapt<2,>=1.10 in /opt/conda/lib/python3.7/site-packages (from deprecated->mapclassify) (1.11.2)
Requirement already satisfied: six>=1.5 in /opt/conda/lib/python3.7/site-packages (from python-dateutil>=2.6.1->pandas->mapclassify) (1.13.0)
In [2]:
# import required libraries

%matplotlib inline
import os
from datetime import datetime

# set environment variable needed for basemap

os.environ["PROJ_LIB"] = r'/opt/conda/pkgs/proj4-5.2.0-he1b5a44_1006/share/proj/'

import numpy as np
import mpl_toolkits

import pandas as pd
import geopandas as gpd
from geopandas import GeoSeries, GeoDataFrame
from shapely.geometry import Point

#import mapclassify
import matplotlib.pyplot as plt
import matplotlib.patches as patches
from mpl_toolkits.basemap import Basemap

import json
import folium
import mplleaflet


Matplotlib is a graphic workhorse for Python and is commonly used for graphs and figures. The %matplotlib inline function helps plots to display properly in Jupyter Notebooks.

In [3]:
%matplotlib inline

Plot a simple function. The ; is used to suppress Matplotlib's written output.

In [4]:
# plot 0 - 9


Basic Plot Framework

A basic Matplotlib plot contains a Figure, Axes, a Title, and Labels.

It is also possible to include multiple axes in one figure. This image shows three axes.

Plotting with Style

Using different plotting character keys in the plot function's second argument, you can plot different styles of lines and symbols using different colors.

In the following plot, '+' indicates the symbols should be shaped like crosses, 'r' tells Matplotlib to plot the line in red, and '--' says the line should be dashed.

In [5]:
plt.plot(range(10), '+r--', markersize=10, label='inc');

Here are a few more examples:

In [6]:
plt.plot(range(10)[::-1], 'b*:', label='dec')
plt.plot([4.5]*10, 'gx-', label='fix')
plt.plot(range(10), 'ko-', label='ver')

Plotting Shapes

You can also draw shapes on a plot using patches.

In [7]:
plt.gca().add_patch(patches.Circle((5, 5), 2, edgecolor='red', facecolor='none'))
plt.gca().add_patch(patches.Rectangle((0, 0), 9, 9, linewidth=10, edgecolor='cyan', facecolor='none'));

A good reference for Matplotlib plotting can be found here.

Plotting with Numpy

Numpy is a Python library that has some useful functions for dealing with number sequences. Below is an example of plotting functions using Matplotlib from a Numpy array.

In [8]:
# evenly sampled time at 200ms intervals
t = np.arange(0., 5., 0.2)

# red dashes, blue squares and green triangles
plt.plot(t, t, 'r--', t, t**2, 'bs', t, t**3, 'g^');

Here is an example using random number generation features.

In [9]:
data = {'a': np.arange(50),
        'c': np.random.randint(0, 50, 50),
        'd': np.random.randn(50)}
data['b'] = data['a'] + 10 * np.random.randn(50)
data['d'] = np.abs(data['d']) * 100

plt.scatter('a', 'b', c='c', s='d', data=data)
plt.xlabel('entry a')
plt.ylabel('entry b');

More Plotting Control

Knowing the Elements of a Plot can allow you to adjust the plot layout with more fine control. Here are important parts of a plot!

Using these we can fine tune labels and line widths for the following histogram.

In [10]:

mu, sigma = 100, 15
x = mu + sigma * np.random.randn(10000)

n, bins, _ = plt.hist(x, 50, density=1, facecolor='g', alpha=0.75)
plt.xlabel('Value', fontsize=22)
plt.ylabel('Probability', fontsize=22)
plt.title('Histogram', fontsize=22)
plt.text(60, .025, r'$\mu=100,\ \sigma=15$', fontsize=22)
plt.axis([40, 160, 0, 0.03])

Introducing Basemap

Basemap is a library for plotting maps in Python. It handles dealing with coordinate projections, plots user-specified data using Matplotlib, and gathers and clips datasets to draw in the background.

Once you set up a Basemap, you can call different functions, such as the drawcoastlines function to add layers to the map.

The cell below initiates a Basemap by designating projection, resolution, window extent and coordinates.

In [11]:
# set initial values

bmap = Basemap(width=12000000,height=9000000,projection='lcc',

Nothing is plotted yet. We can see the type of bmap below.

In [12]:
<mpl_toolkits.basemap.Basemap at 0x7fc5e7a18f98>

Now we can start adding to the map:

In [13]:
# draw coastlines


By adding a blue background, we can simulate oceans.

In [14]:
# set up map

bmap = Basemap(width=12000000,height=9000000,projection='lcc',

# set map background to blue


Finally we can color the continents so that it looks like they are on top of the ocean.

In [15]:
# set up map

bmap = Basemap(width=12000000,height=9000000,projection='lcc',

# fill continents, set lake color same as ocean color.


Basemap Backgrounds

Several types of preloaded map style options are available in Basemap. Here are a few examples:

Blue Marble:

In [16]:
bmap = Basemap(width=12000000,height=9000000,projection='lcc',
Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).

Shaded Relief:

In [17]:
bmap = Basemap(width=12000000,height=9000000,projection='lcc',
Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).

ETOPO1 Global Relief Model:

In [18]:
bmap = Basemap(width=12000000,height=9000000,projection='lcc',
Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).

Projecting with Basemap

Basemap handles different projections with the projection parameter in the basic basemap. It is fairly easy to draw parallels and meridians with Basemap.

This first map has a Lambert Conformal projection, indicated by lcc.

In [19]:
# setup Lambert Conformal basemap

m = Basemap(width=12000000,height=9000000,projection='lcc',