Download notebook (.ipynb)

Visualizing Spatial Information - California Housing#

This demo shows a simple workflow when working with geospatial data:

  • Obtaining a dataset which includes geospatial references.

  • Obtaining a desired geometries (boundaries etc.)

  • Visualisation

In this example we will make a simple proportional symbols map using the California Housing dataset in sklearn package.

import numpy as np
import pandas as pd
import geopandas as gpd

from lets_plot import *
from lets_plot import tilesets

LetsPlot.setup_html()

Prepare the dataset#

from sklearn.datasets import fetch_california_housing

california_housing_bunch = fetch_california_housing()
data = pd.DataFrame(california_housing_bunch.data, columns=california_housing_bunch.feature_names)

# Add $-value field to the dataframe.
# dataset.target: numpy array of shape (20640,)
# Each value corresponds to the average house value in units of 100,000.
data['Value($)'] = california_housing_bunch.target * 100000
data.head()
MedInc HouseAge AveRooms AveBedrms Population AveOccup Latitude Longitude Value($)
0 8.3252 41.0 6.984127 1.023810 322.0 2.555556 37.88 -122.23 452600.0
1 8.3014 21.0 6.238137 0.971880 2401.0 2.109842 37.86 -122.22 358500.0
2 7.2574 52.0 8.288136 1.073446 496.0 2.802260 37.85 -122.24 352100.0
3 5.6431 52.0 5.817352 1.073059 558.0 2.547945 37.85 -122.25 341300.0
4 3.8462 52.0 6.281853 1.081081 565.0 2.181467 37.85 -122.25 342200.0 
# Draw a random sample from the data set.
data = data.sample(n=1000, random_state=42)

Static map#

Let’s create a static map using regular ggplot2 geometries.

Various shape files related to the state of California are available at https://data.ca.gov web site.

For the purpose of this demo the Calofornia State Boundaty zip was downloaded from https://data.ca.gov/dataset/ca-geographic-boundaries and unpacked to ca-state-boundary subdirectory.

Use geopandas to read a shape file to GeoDataFrame#

#CA = gpd.read_file("./ca-state-boundary/CA_State_TIGER2016.shp")

from lets_plot.geo_data import *

CA = geocode_states('CA').scope('US').inc_res(2).get_boundaries()
CA.head()

The geodata is provided by © OpenStreetMap contributors and is made available here under the Open Database License (ODbL).
state found name geometry
0 CA California MULTIPOLYGON (((-124.32694 40.61620, -124.3118...

Keeping in mind that our target is the housing value, fill the choropleth over the state contours using geom_map()function

Make a plot out of polygon and points#

The color of the points will reflect the house age and the size of the points will reflect the value of the house.

# The plot base 
p = ggplot() + scale_color_gradient(name='House Age', low='red', high='green')

# The points layer
points = geom_point(aes(x='Longitude',
                        y='Latitude',
                        size='Value($)',
                        color='HouseAge'), 
                    data=data,
                    alpha=0.8)

# The map
p + geom_polygon(data=CA, fill='#F8F4F0', color='#B71234')\
  + points\
  + theme_void()\
  + ggsize(600, 500)

Interactive map#

The geom_livemap() function creates an interactive base-map super-layer to which other geometry layers are added.

Configuring map tiles#

By default Lets-PLot offers high quality vector map tiles but also can fetch raster tiles from a 3d-party Z-X-Y tile servers.

For the sake of the demo lets use CARTO Voyager tiles by CARTO as our basemap.

LetsPlot.set(tilesets.CARTO_VOYAGER_HIRES)

Make a plot similar to the one above but interactive#

p + geom_livemap()\
  + geom_polygon(data=CA, fill='white', color='#B71234', alpha=0.5)\
  + points

Adjust the initial viewport#

Use parameters location and zoom to define the initial viewport.

# Pass `[lon,lat]` value to the `location` (near Los Angeles)
p + geom_livemap(location=[-118.15, 33.96], zoom=7)\
  + geom_polygon(data=CA, fill='white', color='#B71234', alpha=0.5, size=1)\
  + points