Visualize results with rt.plot()

rt.plot() is a lightweight plotting method included with GeoAnalytics for Microsoft Fabric that allows you to quickly view raster data stored in a PySpark DataFrame. This tutorial will show you how to use rt.plot along with various parameters to plot raster data, symbolize with a color map, add a basemap and use a specified extent and spatial reference. The full list of supported parameters is documented in the API reference for rt.plot.

Steps

Import

In your notebook, import geoanalytics_fabric. Also, import the modules required to run the plotting examples below.

1
2
3
4
5
6
7

import geoanalytics_fabric

import geoanalytics_fabric.sql.functions as ST
import pyspark.sql.functions as F
from geoanalytics_fabric.raster import functions as RT
from geoanalytics_fabric.tools.raster import *

Create a PySpark DataFrame with a raster column

1. Load a polygon feature service of earthquakes since 1970 into a DataFrame.

   Python

   Use dark colors for code blocksCopy

   1
   2
   3
   4
   5
   6
   7
   8
   
   # Path to the earthquakes data
   data_path = r"https://sampleserver6.arcgisonline.com/arcgis/rest/services/Earthquakes_Since1970/FeatureServer/0"
   
   earthquakes_df = spark.read.format("feature-service").load(data_path)
   
   # Create a DataFrame for earthquakes in Greece using ST_EnvIntersects
   greece_earthquakes_df = earthquakes_df.where(ST.bbox_intersects("shape", xmin=19.28, xmax=26.65, ymin=35.13, ymax=40.17))

2. Create square bins for each earthquake and group the bins by count.

   Python

   Use dark colors for code blocksCopy

   1
   2
   3
   4
   5
   6
   
   # Create the bins
   bins = greece_earthquakes_df\
     .withColumn("bin", ST.square_bin("shape", 1))\
     .groupBy("bin").count()\
     .withColumn("count", F.col("count").cast("int"))

3. Run the BinsToRaster tool to create a raster from these bins.

   Python

   Use dark colors for code blocksCopy

   1
   2
   3
   4
   5
   6
   7
   
   # Use the BinsToRaster tool to convert bins to raster
   raster = BinsToRaster()\
     .setBinColumn("bin")\
     .setBandValueColumns("count")\
     .setTileSize(10, 10)\
     .run(bins)

Plot the raster data

In this example the figsize parameter is used. This parameter adjusts the size of the result plot.

1
2

raster.rt.plot(figsize=(6, 6))

Plot with a basemap and extent

In this example the following parameters are introduced:

• basemap—Adds a basemap to the plot. Choose from "light" (Light Gray Canvas), "dark" (Dark Gray Canvas), "streets" (Esri Streets Basemap) or "osm" (OpenStreetMap Vector Basemap).
• sr—The spatial reference used to plot the data.
• extent—Limits the extent of the plot. The desired plot extent is defined using a tuple in the format (min_x, min_y, max_x, max_y).

1
2
3

margin=2
raster.rt.plot(figsize=(6, 6), basemap="streets", sr=4326, extent=(19.28 - margin, 35.13 - margin, 26.65 + margin, 41.17 + margin))

Plot with a color map

In this example the cmap parameter is introduced. This is used to represent your data in 3D colorspace. To learn more about colormaps, see the see the matplotlib's documentation for colormaps

1
2
3

margin=2
raster.rt.plot(figsize=(6, 6), cmap='afmhot', sr=4326, basemap="streets", extent=(19.28 - margin, 35.13 - margin, 26.65 + margin, 41.17 + margin))

Plot with an alpha

As you can see in the previous example, the raster layer is hiding the basemap. In this example the alpha parameter will be used to control the transparency of the plotted elements.

1
2
3

margin=2
raster.rt.plot(figsize=(6, 6), cmap='afmhot', sr=4326, basemap="streets", extent=(19.28 - margin, 35.13 - margin, 26.65 + margin, 41.17 + margin), alpha=0.6)

What's next?

Because rt.plot extends matplotlib, you can leverage the rich plotting capabilities in matplotlib for plotting your raster data. The examples shown here are a basic introduction to what's possible with rt.plot and matplotlib. For more ideas see the matplotlib's documentation.