raster_conditional.py

import numpy as np
import rasterio



# open loss and grab metadata about the tif,
# like it's location / projection / cellsize
with rasterio.open('loss.tif') as loss_src:
    kwargs = loss_src.meta

    # also grab the windows (stripes? tiles?) so we can iterate
    # over the entire tif without running out of memory
    windows = loss_src.block_windows(1)

    # open gain
    with rasterio.open('gain.tif') as gain_src:

        # open extent
        with rasterio.open('extent.tif') as extent_src:

            # update those kwargs for the dataset we'll write out 
            kwargs.update(
                driver='GTiff',
                count=1,
                compress='lzw',
                nodata=0
            )

            with rasterio.open('output.tif', 'w', **kwargs) as dst:
                for idx, window in windows:
                    loss_data = loss_src.read(1, window=window)
                    gain_data = gain_src.read(1, window=window)
                    extent_data = extent_src.read(1, window=window)

                    # create an empty array
                    dst_data = np.zeros((window.height, window.width), dtype='uint8')

                    # where loss & gain, set output to 100, otherwise keep dst_data value
                    dst_data[np.where((loss_data > 0) & (gain_data > 0))] = 100

                    # etc
                    dst_data[np.where((loss_data > 10) & (gain_data > 0) & (extent_data > 50))] = 99
                    dst_data[np.where((loss_data < 10) & (gain_data == 0) & (extent_data < 50))] = 1
                    dst_data[np.where((loss_data > 10) & (gain_data > 0) & (extent_data >= 90))] = 2

                    dst.write_band(1, dst_data, window=window)