Building 2-layer maps - combining polygons and scatterplots¶

In this chapter, you will learn how to create a two-layer map by first plotting regions from a shapefile and then plotting location points as a scatterplot. This is the Summary of lecture "Visualizing Geospatial Data in Python", via datacamp.

toc: true
badges: true
comments: true
author: Chanseok Kang
categories: [Python, Datacamp, Visualization]
image: images/nashville_chicken_permits.png

In [1]:

import pandas as pd
import matplotlib.pyplot as plt

plt.rcParams['figure.figsize'] = (10, 5)

Introduction¶

Styling a scatterplot¶

In this exercise, you'll be using plt.scatter() to plot the father and son height data from the video. The father_son DataFrame is available in your workspace. In each scatterplot, plot father_son.fheight as x-axis and father_son.sheight as y-axis.

In [2]:

father_son = pd.read_csv('./dataset/father_son.csv', usecols=['fheight', 'sheight'])
father_son.head()

Out[2]:

	fheight	sheight
0	65.04851	59.77827
1	63.25094	63.21404
2	64.95532	63.34242
3	65.75250	62.79238
4	61.13723	64.28113

In [3]:

# Scatterplot 1 - father heights vs son heights with darkred square markers
plt.scatter(father_son.fheight, father_son.sheight, c='darkred', marker='s');

In [4]:

# Scatterplot 2 - yellow markers with darkblue borders
plt.scatter(father_son.fheight, father_son.sheight, c = 'yellow', edgecolor= 'darkblue');

In [5]:

# Scatterplot 3
plt.scatter(father_son.fheight, father_son.sheight,  c = 'yellow', edgecolor = 'darkblue');
plt.grid();
plt.xlabel('father height (inches)');
plt.ylabel('son height (inches)');
plt.title('Son Height as a Function of Father Height');

Plotting chicken locations¶

Now you will create a scatterplot that shows where the Nashville chickens are!

In [6]:

# Load the dataset
chickens = pd.read_csv('./dataset/chickens.csv')

# Look at the first few rows of the chickens DataFrame
print(chickens.head())

# Plot the locations of all Nashville chicken permits
plt.scatter(x=chickens.lng, y=chickens.lat);

   Unnamed: 0  permit       city state    zip  district date_issued  \
0           0   288.0  Nashville    TN  37211      30.0  11/26/2014   
1           1   435.0  Nashville    TN  37211      19.0  05/18/2015   
2           2   270.0  Nashville    TN  37216       7.0  11/20/2014   
3           3   249.0  Nashville    TN  37216      25.0  12/29/2014   
4           4   200.0  Nashville    TN  37221      22.0  12/10/2014   

         lat        lng  
0  36.029132 -86.724578  
1  36.069132 -86.754578  
2  36.236134 -86.711204  
3  36.246134 -86.751204  
4  36.029367 -86.935290

We can see the relative location of the Nashville chickens using longitude for the x-axis and latitude for the y-axis. Next, now you'll up your game by plotting the locations on an actual map!

Geometries and shapefiles¶

Shapefiles
- stores a special type of data known as geometry
  - Point
  - Line
  - Polygon
- .shp : contains the geometry
- .dbf : holds attributes for each geometry
- .shx : links the attributes to the geometry

Creating a GeoDataFrame & examining the geometry¶

Let's see where service districts are in Nashville.

Note: if you use anaconda on windows, you must be faced in GDAL error. At that case, follow the instruction here

In [12]:

import geopandas as gpd

# Read in the services district shapefile and look at the first few rows
service_district = gpd.read_file('./dataset/ServiceDistricts.shp')
print(service_district.head())

# Print the contents of the service districts geometry in the first row
# print(service_district.loc[0, 'geometry'])

   OBJECTID  AREA_SQ_MI                       Name  \
0         0         198    Urban Services District   
1         4         327  General Services District   

                                            geometry  
0  POLYGON ((1766078.549 711757.270, 1766000.028 ...  
1  MULTIPOLYGON (((1800630.485 619627.187, 180060...

The geometry field holds a series of latitude/longitude pairs that define the border of a polygon. Printing a geometry field gives all of the latitude/longitude pairs that make up the polygon boundary.

Plotting shapefile polygons¶

The next step is to show the map of polygons.

In [8]:

# Plot the Service Districts without any additional arguments
service_district.to_crs(epsg=4326).plot(figsize=(10, 10));

# Plot the Service Districts, color them according to name, and show a legend
service_district.to_crs(epsg=4326).plot(column='Name', legend=True, figsize=(10, 10));

Scatterplots over polygons¶

Plotting points over polygons - part 1¶

Make a basic plot of the service districts with the chicken locations. The packages needed have already been imported for you. The chickens DataFrame and service_district GeoDataFrame are in your workspace.

In [9]:

# Plot the service district shapefile
service_district.to_crs(epsg=4326).plot(column='Name', figsize=(10, 10));

# Add the chicken locations
plt.scatter(x=chickens.lng, y=chickens.lat, c='black');

Plotting points over polygons - part 2¶

Plot the service districts and chicken permits together to see what story your visualization tells.

In [11]:

# Plot the service district shapefile
service_district.to_crs(epsg=4326).plot(column='Name', legend=True);

# Add the chicken locations
plt.scatter(x=chickens.lng, y=chickens.lat, c='black', edgecolor='white');

# Add labels and title
plt.title('Nashville Chicken Permits');
plt.xlabel('longitude');
plt.ylabel('latitude');

# Add grid lines
plt.grid();