Using geocube in Python to rasterize each polygon in shapefile

Question

I have a shapefile of polygons that I brought into Python in Jupyter Notebook via GeoPandas as a geodataframe. There are hundreds of polygons in this shapefile, and I want to create an individual raster for each polygon, where each raster has that rasterized polygon isolated, as if it were the only polygon, where each output raster has a unique number attached to it (e.g. "Output_raster_35" etc.). I want to specifically burn in values corresponding to each polygon within the "Values" column in this geodataframe.

I am trying to use the geocube package for this task by running the following to rasterize the whole shapefile:

vector_fn = 'polygons.shp'

out_grid = make_geocube(
    vector_data=vector_fn,
    measurements=["Values"],
    resolution=(-25, 25),
    fill=-9999,
)
out_grid["Values"].rio.to_raster("my_rasterized_column.tif")

This code was suggested from this stackoverflow post: Rasterizing polygons in shapefile using GDAL

However, I need to iterate this process over each polygon (row), rather than rasterizing all polygons at once.

So far my idea is:

polygons = gpd.read_file('polygons.shp')
polygon_IDs = polygons.Polygon_ID.tolist()

for i in polygon_IDs:
    x = polygons.loc[polygons.Polygon_ID == i]
    vector_fn = x

    out_grid = make_geocube(
        vector_data=vector_fn,
        measurements=["Values"],
        resolution=(-25, 25),
        fill=-9999,
    )
    out_grid["Values"].rio.to_raster("Output_Raster")

Though I am having trouble sorting out whether I have my logic and syntax correct here.

How can I fix my code so that it creates individual burned rasters for each polygon in the shapefile where output raster files are numbered?

Answer 1

I changed a couple of things:

Replaced polygon_ID with ['id']. That is how you call a column/attribute from a dataframe.

Added str(i) to your saving section to save the next polygon that is selected through your for loop.

import geopandas as gpd
from geocube.api.core import make_geocube

polygons = gpd.read_file('testing.shp')
polygon_IDs = polygons['id'].tolist()

for i in polygon_IDs:
    x = polygons.loc[polygons['id'] == i]
    vector_fn = x
    out_grid = make_geocube(
        vector_data=vector_fn,
        measurements=["values"],
        resolution=(-25, 25),
        fill=-9999,
    )
    out_grid["values"].rio.to_raster(str(i) + "_Output_Raster.tif")