Oracle Spatial AI implements a plotting functionality for clusters in the plot_clusters function. The following code passes the training data and the labels as parameters. The with_noise=False parameter avoids displaying the observations labeled with –1. The with_basemap=True parameter sets a basemap as background.

from oraclesai.vis import plot_clusters

fig, ax = plt.subplots(figsize=(12,12))  

plot_clusters(X, lisa_labels, with_noise=False, with_basemap=True, cmap='Dark2', ax=ax)

The result consists of those observations with a local Moran’s I statistically significant, colored according to their quadrant. Quadrants are defined as follows:

1. A high value surrounded by high values (hot spots).
2. A low value around high values (outliers).
3. A low value surrounded by low values (cold spots).
4. A high value around high values (outliers).