Benjamin Weinstein

I'm still debugging, i'll update when I fix. ``` from functools import partial from rasterio.enums import MergeAlg from geocube.rasterize import rasterize_image g = gpd.read_file("/Users/ben/Dropbox/Weecology/Crowns/examples/2019_BART_5_320000_4881000_image.shp") g["mask"]=1 cube = make_geocube(vector_data=g, resolution=(100, 100),...

I wasn't sure where to place, but tried both, no change. ``` g = gpd.read_file("/Users/ben/Dropbox/Weecology/Crowns/examples/2019_BART_5_320000_4881000_image.shp") g["mask"]=1 cube = make_geocube(vector_data=g, resolution=(100, 100), measurements=['mask'] , rasterize_function=partial(rasterize_image, merge_alg=MergeAlg.add, fill=0)) cube.plot() cube.rio.to_raster("/Users/Ben/Desktop/data1.tif") ``` ```...

yup. that did it. I'm away for the weekend, but i'll try to fork and submit a notebook PR to add as a resource on monday. Thanks. On Thu, Jun...

Is there suggested best practice for dask-jobqueue for flushing worker logs on SLURM. I'm finding that i'm not getting print statements until the process is tearing down, sometimes hours later....

What would it take to move to tf 2.0? Any known limitation? Just wondering. integration with Keras is always attractive.

@AakashKumarNain I have a similar use case, did you find performance improvements when transitioning from Numpy to dask, reading image slices from file?

@mrocklin I think I'm stumbling on the exact issue "#maybe bad" mentioned at the top. psuedo code (working on reproducible) We have a large number of numpy arrays (geospatial tiles)...

I created a working example here for those who find this link: https://github.com/dask/distributed/issues/2333

@mrocklin happy to help if I can with tests, I use this kind of workflow frequently with dask-jobqueue for submitting to GPU clusters On Sat, May 23, 2020 at 10:16...

@mrts Did this solve your problem, i'm having the same issue with git bash.