Improvements under the hood

TIF LZW Compression

The output files can take up a lot of storage space. Fortunately, GDAL directly offers an option to compress TIF files with the lossless LZW compression algorithm. This is the new default setting for TIF output:

dataset = driver.Create(
     outpath,
     y_pixels,
     x_pixels,
     1,
     gdal.GDT_Float32,
     options=['COMPRESS=LZW'])
 dataset.SetGeoTransform(self.geo_transform)
 dataset.SetProjection(self.wkt_projection)

Automatic Thread Count

The number of threads for computation can be set automatically depending on the available CPU. The optimal thread count was tested on various systems with 12, 24, and 64 cores. Beyond 16 threads, no improved speeds were achieved, so the number of threads is capped at 16:

self.threads = min(os.cpu_count(), 16)

Structural Adjustments

Under the hood, there are a few adjustments. The essential parameters can now be set directly from main.py, so you no longer need to go into the manager.py and model.py files for this.