Closed mmann1123 closed 2 years ago
The problem is not with n_chunks
but likely with n_workers
. What OS are you running this on?
If you do:
with gw.open(l8_224078_20200518) as src:
# Write the data to a GeoTiff
src.sel(band=[3, 2, 1]).gw.to_raster('./output/output.tif',
n_workers=1, n_threads=4, n_chunks=16,
overwrite=True)
does it fix the block issue?
Yes that resolves it.
This might be my lack of understanding, but
n_chunks
seems to create problems when writing withn_worker
andn_threads
set.