# \`save\_cog\_with\_dask\`: Cannot convert fill\_value 999999 to dtype uint8

**URL:** <https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713>\
**Category:** Science\
**Created:** [November 22, 2024, 3:19pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713 "2024-11-22T15:19:29Z")\
**Posts on this page:** 10\
**Page:** 1

<div class="post-metadata">

**Author:** ![remi-braun](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/remi-braun/32/1178_2.png) [@remi-braun](https://discourse.pangeo.io/u/remi-braun)\
**Post date:** [November 22, 2024, 3:19pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/1 "2024-11-22T15:19:29Z")

</div>

Hi there 👋,

I am currently trying to write COGs via Dask with `odc.geo.cog.save_cog_with_dask`.  
Most of the time, it seems fine, but with the same code, deserialisation issue happen sometimes:

```auto
  File "C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py", line 489, in _check_fill_value
    fill_value = np.asarray(fill_value, dtype=ndtype)
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
OverflowError: Python integer 999999 out of bounds for uint8

```

My code is:

```auto
cog.save_cog_with_dask(
    xds.copy(data=xds.fillna(nodata).astype(dtype)).rio.set_nodata(nodata),
    str(path),
).compute()

```

I triple checked, everything is fine in the nodata area, for both `rioxarray` and `odc` accessors and encoding.0 in float shouldn’t be an issue as they are safely cast into uint8 data.

 ![2024-11-22_09h48_39](https://canada1.discourse-cdn.com/flex030/uploads/pangeo/original/2X/3/34a80cfa5871ec6b1cdc6d70a6395b2fcf0c359e.png)

> **Full stack trace**
>
> 2024-11-22 09:17:34,307 - distributed.protocol.core - CRITICAL - Failed to deserialize  
> Traceback (most recent call last):  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py”, line 489, in \_check\_fill\_value  
> fill\_value = np.asarray(fill\_value, dtype=ndtype)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> OverflowError: Python integer 999999 out of bounds for uint8  
> The above exception was the direct cause of the following exception:  
> Traceback (most recent call last):  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\core.py”, line 175, in loads  
> return msgpack.loads(  
> ^^^^^^^^^^^^^^  
> File “msgpack\_unpacker.pyx”, line 194, in msgpack.\_cmsgpack.unpackb  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\core.py”, line 159, in \_decode\_default  
> return merge\_and\_deserialize(  
> ^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\contextlib.py”, line 81, in inner  
> return func(\*args, \*\*kwds)  
> ^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\serialize.py”, line 525, in merge\_and\_deserialize  
> return deserialize(header, merged\_frames, deserializers=deserializers)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\serialize.py”, line 452, in deserialize  
> return loads(header, frames)  
> ^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\serialize.py”, line 66, in dask\_loads  
> return loads(header[“sub-header”], frames)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\numpy.py”, line 219, in deserialize\_numpy\_maskedarray  
> return np.ma.masked\_array(data, mask=mask, fill\_value=fill\_value)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py”, line 2968, in **new**  
> \_data.\_fill\_value = \_check\_fill\_value(fill\_value, \_data.dtype)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py”, line 495, in \_check\_fill\_value  
> raise TypeError(err\_msg % (fill\_value, ndtype)) from e  
> TypeError: Cannot convert fill\_value 999999 to dtype uint8  
> 2024-11-22 09:17:34,312 - distributed.worker - ERROR - Cannot convert fill\_value 999999 to dtype uint8  
> Traceback (most recent call last):  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py”, line 489, in \_check\_fill\_value  
> fill\_value = np.asarray(fill\_value, dtype=ndtype)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> OverflowError: Python integer 999999 out of bounds for uint8  
> The above exception was the direct cause of the following exception:  
> Traceback (most recent call last):  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\worker.py”, line 2075, in gather\_dep  
> response = await get\_data\_from\_worker(  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\worker.py”, line 2881, in get\_data\_from\_worker  
> response = await send\_recv(  
> ^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\core.py”, line 1018, in send\_recv  
> response = await comm.read(deserializers=deserializers)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\comm\tcp.py”, line 247, in read  
> msg = await from\_frames(  
> ^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\comm\utils.py”, line 78, in from\_frames  
> res = \_from\_frames()  
> ^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\comm\utils.py”, line 61, in \_from\_frames  
> return protocol.loads(  
> ^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\core.py”, line 175, in loads  
> return msgpack.loads(  
> ^^^^^^^^^^^^^^  
> File “msgpack\_unpacker.pyx”, line 194, in msgpack.\_cmsgpack.unpackb  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\core.py”, line 159, in \_decode\_default  
> return merge\_and\_deserialize(  
> ^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\contextlib.py”, line 81, in inner  
> return func(\*args, \*\*kwds)  
> ^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\serialize.py”, line 525, in merge\_and\_deserialize  
> return deserialize(header, merged\_frames, deserializers=deserializers)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\serialize.py”, line 452, in deserialize  
> return loads(header, frames)  
> ^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\serialize.py”, line 66, in dask\_loads  
> return loads(header[“sub-header”], frames)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\protocol\numpy.py”, line 219, in deserialize\_numpy\_maskedarray  
> return np.ma.masked\_array(data, mask=mask, fill\_value=fill\_value)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py”, line 2968, in **new**  
> \_data.\_fill\_value = \_check\_fill\_value(fill\_value, \_data.dtype)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py”, line 495, in \_check\_fill\_value  
> raise TypeError(err\_msg % (fill\_value, ndtype)) from e  
> TypeError: Cannot convert fill\_value 999999 to dtype uint8  
> 2024-11-22 09:17:34,367 - distributed.core - ERROR - Exception while handling op get\_data  
> Traceback (most recent call last):  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\core.py”, line 834, in \_handle\_comm  
> result = await result  
> ^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\core.py”, line 981, in wrapper  
> return await func(self, \*args, \*\*kwargs)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\worker.py”, line 1798, in get\_data  
> assert response == “OK”, response  
> ^^^^^^^^^^^^^^^^  
> AssertionError: {‘op’: ‘get\_data’, ‘keys’: {(‘from\_sequence-9bfcd20b0840146be3c99f402b3807dc’, 83)}, ‘who’: ‘tcp://127.0.0.1:57260’, ‘reply’: True}  
> 2024-11-22 09:17:34,372 - distributed.worker - ERROR - {‘op’: ‘get\_data’, ‘keys’: {(‘from\_sequence-9bfcd20b0840146be3c99f402b3807dc’, 83)}, ‘who’: ‘tcp://127.0.0.1:57260’, ‘reply’: True}  
> Traceback (most recent call last):  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\worker.py”, line 2075, in gather\_dep  
> response = await get\_data\_from\_worker(  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\worker.py”, line 2881, in get\_data\_from\_worker  
> response = await send\_recv(  
> ^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\core.py”, line 1043, in send\_recv  
> raise exc.with\_traceback(tb)  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\core.py”, line 834, in \_handle\_comm  
> result = await result  
> ^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\core.py”, line 981, in wrapper  
> return await func(self, \*args, \*\*kwargs)  
> ^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\distributed\worker.py”, line 1798, in get\_data  
> assert response == “OK”, response  
> ^^^^^^^^^^^^^^^^^  
> AssertionError: {‘op’: ‘get\_data’, ‘keys’: {(‘from\_sequence-9bfcd20b0840146be3c99f402b3807dc’, 83)}, ‘who’: ‘tcp://127.0.0.1:57260’, ‘reply’: True}

Does somebody ever encountered something like that?

Sorry it’s difficult to adress a minimum workable example as the rasters I’m using are very heavy (and commercial data).

> **Conda list**
>
> conda list
> 
> # packages in environment at C:\Users\rbraun\Anaconda3\envs\eoprocesses:
> 
> # 
> 
> # Name Version Build Channel
> 
> affine 2.4.0 pypi\_0 pypi  
> art 6.2 pypi\_0 pypi  
> asciitree 0.3.3 pypi\_0 pypi  
> astroid 3.2.4 pypi\_0 pypi  
> attrs 23.2.0 pypi\_0 pypi  
> azure-core 1.30.2 pypi\_0 pypi  
> azure-storage-blob 12.21.0 pypi\_0 pypi  
> backports-tarfile 1.2.0 pypi\_0 pypi  
> beautifulsoup4 4.12.3 pypi\_0 pypi  
> black 24.8.0 pypi\_0 pypi  
> bokeh 3.5.1 pypi\_0 pypi  
> boto3 1.34.149 pypi\_0 pypi  
> botocore 1.34.149 pypi\_0 pypi  
> bzip2 1.0.8 h2466b09\_7 conda-forge  
> ca-certificates 2024.8.30 h56e8100\_0 conda-forge  
> cachetools 5.4.0 pypi\_0 pypi  
> certifi 2024.7.4 pypi\_0 pypi  
> cffi 1.16.0 pypi\_0 pypi  
> cfgv 3.4.0 pypi\_0 pypi  
> charset-normalizer 3.3.2 pypi\_0 pypi  
> click 8.1.7 pypi\_0 pypi  
> click-plugins 1.1.1 pypi\_0 pypi  
> cligj 0.7.2 pypi\_0 pypi  
> cloudpathlib 0.18.1 pypi\_0 pypi  
> cloudpickle 3.0.0 pypi\_0 pypi  
> color-operations 0.1.6 pypi\_0 pypi  
> colorama 0.4.6 pypi\_0 pypi  
> colorlog 6.8.2 pypi\_0 pypi  
> configobj 5.0.8 pypi\_0 pypi  
> contourpy 1.2.1 pypi\_0 pypi  
> coverage 7.6.1 pypi\_0 pypi  
> cryptography 43.0.0 pypi\_0 pypi  
> cycler 0.12.1 pypi\_0 pypi  
> daal 2024.5.0 pypi\_0 pypi  
> daal4py 2024.5.0 pypi\_0 pypi  
> dask 2024.11.2 pypi\_0 pypi  
> dask-expr 1.1.19 pypi\_0 pypi  
> dicttoxml 1.7.16 pypi\_0 pypi  
> dill 0.3.8 pypi\_0 pypi  
> distlib 0.3.8 pypi\_0 pypi  
> distributed 2024.11.2 pypi\_0 pypi  
> docutils 0.21.2 pypi\_0 pypi  
> donfig 0.8.1.post1 pypi\_0 pypi  
> earthengine-api 0.1.413 pypi\_0 pypi  
> ee-extra 0.0.15 pypi\_0 pypi  
> eemont 0.3.6 pypi\_0 pypi  
> eoreader 0.21.7 pypi\_0 pypi  
> eosets 0.2.5 pypi\_0 pypi  
> fasteners 0.19 pypi\_0 pypi  
> filelock 3.15.4 pypi\_0 pypi  
> flake8 7.1.1 pypi\_0 pypi  
> fonttools 4.53.1 pypi\_0 pypi  
> fsspec 2024.6.1 pypi\_0 pypi  
> geographiclib 2.0 pypi\_0 pypi  
> geopandas 1.0.1 pypi\_0 pypi  
> geopy 2.4.1 pypi\_0 pypi  
> google-api-core 2.19.1 pypi\_0 pypi  
> google-api-python-client 2.138.0 pypi\_0 pypi  
> google-auth 2.32.0 pypi\_0 pypi  
> google-auth-httplib2 0.2.0 pypi\_0 pypi  
> google-cloud-core 2.4.1 pypi\_0 pypi  
> google-cloud-storage 2.18.0 pypi\_0 pypi  
> google-crc32c 1.5.0 pypi\_0 pypi  
> google-resumable-media 2.7.1 pypi\_0 pypi  
> googleapis-common-protos 1.63.2 pypi\_0 pypi  
> h5netcdf 1.3.0 pypi\_0 pypi  
> h5py 3.11.0 pypi\_0 pypi  
> httplib2 0.22.0 pypi\_0 pypi  
> identify 2.6.0 pypi\_0 pypi  
> idna 3.7 pypi\_0 pypi  
> imageio 2.34.2 pypi\_0 pypi  
> importlib-metadata 8.2.0 pypi\_0 pypi  
> iniconfig 2.0.0 pypi\_0 pypi  
> isodate 0.6.1 pypi\_0 pypi  
> isort 5.13.2 pypi\_0 pypi  
> jaraco-classes 3.4.0 pypi\_0 pypi  
> jaraco-context 5.3.0 pypi\_0 pypi  
> jaraco-functools 4.0.1 pypi\_0 pypi  
> jinja2 3.1.4 pypi\_0 pypi  
> jmespath 1.0.1 pypi\_0 pypi  
> joblib 1.4.2 pypi\_0 pypi  
> jsonschema 4.23.0 pypi\_0 pypi  
> jsonschema-specifications 2023.12.1 pypi\_0 pypi  
> keyring 25.2.1 pypi\_0 pypi  
> kiwisolver 1.4.5 pypi\_0 pypi  
> lazy-loader 0.4 pypi\_0 pypi  
> libexpat 2.6.4 he0c23c2\_0 conda-forge  
> libffi 3.4.2 h8ffe710\_5 conda-forge  
> libsqlite 3.47.0 h2466b09\_1 conda-forge  
> libzlib 1.3.1 h2466b09\_1 conda-forge  
> locket 1.0.0 pypi\_0 pypi  
> lxml 5.3.0 pypi\_0 pypi  
> lz4 4.3.3 pypi\_0 pypi  
> markdown-it-py 3.0.0 pypi\_0 pypi  
> markupsafe 2.1.5 pypi\_0 pypi  
> matplotlib 3.9.2 pypi\_0 pypi  
> mccabe 0.7.0 pypi\_0 pypi  
> mdurl 0.1.2 pypi\_0 pypi  
> methodtools 0.4.7 pypi\_0 pypi  
> more-itertools 10.3.0 pypi\_0 pypi  
> msgpack 1.0.8 pypi\_0 pypi  
> mypy-extensions 1.0.0 pypi\_0 pypi  
> networkx 3.3 pypi\_0 pypi  
> nh3 0.2.18 pypi\_0 pypi  
> nodeenv 1.9.1 pypi\_0 pypi  
> numcodecs 0.13.0 pypi\_0 pypi  
> numpy 2.0.1 pypi\_0 pypi  
> odc-geo 0.4.8 pypi\_0 pypi  
> openssl 3.4.0 h2466b09\_0 conda-forge  
> packaging 24.1 pypi\_0 pypi  
> pandas 2.2.2 pypi\_0 pypi  
> partd 1.4.2 pypi\_0 pypi  
> pathspec 0.12.1 pypi\_0 pypi  
> pillow 10.4.0 pypi\_0 pypi  
> pip 24.3.1 pyh8b19718\_0 conda-forge  
> pkginfo 1.10.0 pypi\_0 pypi  
> platformdirs 4.2.2 pypi\_0 pypi  
> pluggy 1.5.0 pypi\_0 pypi  
> pre-commit 3.8.0 pypi\_0 pypi  
> proto-plus 1.24.0 pypi\_0 pypi  
> protobuf 5.27.2 pypi\_0 pypi  
> psutil 6.0.0 pypi\_0 pypi  
> pyarrow 17.0.0 pypi\_0 pypi  
> pyarrow-hotfix 0.6 pypi\_0 pypi  
> pyasn1 0.6.0 pypi\_0 pypi  
> pyasn1-modules 0.4.0 pypi\_0 pypi  
> pycm 4.0 pypi\_0 pypi  
> pycodestyle 2.12.0 pypi\_0 pypi  
> pycparser 2.22 pypi\_0 pypi  
> pyflakes 3.2.0 pypi\_0 pypi  
> pygments 2.18.0 pypi\_0 pypi  
> pykdtree 1.3.12 pypi\_0 pypi  
> pylint 3.2.6 pypi\_0 pypi  
> pyogrio 0.9.0 pypi\_0 pypi  
> pyparsing 3.1.2 pypi\_0 pypi  
> pyproj 3.6.1 pypi\_0 pypi  
> pyresample 1.29.0 pypi\_0 pypi  
> pystac 1.10.1 pypi\_0 pypi  
> pytest 8.3.2 pypi\_0 pypi  
> pytest-cov 5.0.0 pypi\_0 pypi  
> python 3.11.10 hce54a09\_3\_cpython conda-forge  
> python-box 7.2.0 pypi\_0 pypi  
> python-dateutil 2.9.0.post0 pypi\_0 pypi  
> pytz 2024.1 pypi\_0 pypi  
> pywin32-ctypes 0.2.2 pypi\_0 pypi  
> pyyaml 6.0.1 pypi\_0 pypi  
> rasterio 1.3.10 pypi\_0 pypi  
> readme-renderer 44.0 pypi\_0 pypi  
> referencing 0.35.1 pypi\_0 pypi  
> requests 2.32.3 pypi\_0 pypi  
> requests-toolbelt 1.0.0 pypi\_0 pypi  
> rfc3986 2.0.0 pypi\_0 pypi  
> rich 13.7.1 pypi\_0 pypi  
> rioxarray 0.17.0 pypi\_0 pypi  
> rpds-py 0.19.1 pypi\_0 pypi  
> rsa 4.9 pypi\_0 pypi  
> rtree 1.3.0 pypi\_0 pypi  
> s3transfer 0.10.2 pypi\_0 pypi  
> scikit-image 0.24.0 pypi\_0 pypi  
> scikit-learn 1.5.1 pypi\_0 pypi  
> scikit-learn-intelex 2024.5.0 pypi\_0 pypi  
> scipy 1.14.0 pypi\_0 pypi  
> seaborn 0.13.2 pypi\_0 pypi  
> sertit 1.43.1.dev6 pypi\_0 pypi  
> sertit-models 0.4.1 pypi\_0 pypi  
> setuptools 71.0.4 pyhd8ed1ab\_0 conda-forge  
> shapely 2.0.5 pypi\_0 pypi  
> six 1.16.0 pypi\_0 pypi  
> snuggs 1.4.7 pypi\_0 pypi  
> sortedcontainers 2.4.0 pypi\_0 pypi  
> soupsieve 2.5 pypi\_0 pypi  
> spyndex 0.6.0 pypi\_0 pypi  
> tbb 2021.13.0 pypi\_0 pypi  
> tblib 3.0.0 pypi\_0 pypi  
> tempenv 2.0.0 pypi\_0 pypi  
> threadpoolctl 3.5.0 pypi\_0 pypi  
> tifffile 2024.7.24 pypi\_0 pypi  
> tk 8.6.13 h5226925\_1 conda-forge  
> tomlkit 0.13.0 pypi\_0 pypi  
> toolz 0.12.1 pypi\_0 pypi  
> tornado 6.4.1 pypi\_0 pypi  
> tqdm 4.66.4 pypi\_0 pypi  
> twine 5.1.1 pypi\_0 pypi  
> typing-extensions 4.12.2 pypi\_0 pypi  
> tzdata 2024.1 pypi\_0 pypi  
> ucrt 10.0.22621.0 h57928b3\_0 conda-forge  
> uritemplate 4.1.1 pypi\_0 pypi  
> urllib3 2.2.2 pypi\_0 pypi  
> validators 0.33.0 pypi\_0 pypi  
> vc 14.3 h8a93ad2\_20 conda-forge  
> vc14\_runtime 14.40.33810 ha82c5b3\_20 conda-forge  
> virtualenv 20.26.3 pypi\_0 pypi  
> vs2015\_runtime 14.40.33810 h3bf8584\_20 conda-forge  
> wheel 0.43.0 pyhd8ed1ab\_1 conda-forge  
> wirerope 0.4.7 pypi\_0 pypi  
> xarray 2024.7.0 pypi\_0 pypi  
> xgboost 2.1.0 pypi\_0 pypi  
> xyzservices 2024.6.0 pypi\_0 pypi  
> xz 5.2.6 h8d14728\_0 conda-forge  
> zarr 2.18.2 pypi\_0 pypi  
> zict 3.0.0 pypi\_0 pypi  
> zipp 3.19.2 pypi\_0 pypi

NB: It’s a repost from [`save_cog_with_dask`: Cannot convert fill\_value 999999 to dtype uint8 - Dask Forum](https://dask.discourse.group/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/3688)  
I don’t think (or know) if it’s a bug, sio I rather start the discussion here before creating an issue there.

---

<div class="post-metadata">

**Author:** ![chuckwondo](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/chuckwondo/32/2170_2.png) [@chuckwondo](https://discourse.pangeo.io/u/chuckwondo)\
**Post date:** [November 23, 2024, 12:21pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/2 "2024-11-23T12:21:25Z")

</div>

It appears that you are attempting to create an array of type `uint8`, which can only hold integer values in the range [0, 255], but attempting to insert the value `999999`, which is far outside of that range. You either need to choose a value within that range, or use a dtype with a range that includes your large value.

---

<div class="post-metadata">

**Author:** ![remi-braun](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/remi-braun/32/1178_2.png) [@remi-braun](https://discourse.pangeo.io/u/remi-braun)\
**Post date:** [November 23, 2024, 2:43pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/3 "2024-11-23T14:43:59Z")

</div>

This is where I don’t understand, where does this value 999999 comes from?  
You can see on the screenshot that all the nodata values are set to 0.

---

<div class="post-metadata">

**Author:** ![rabernat](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/rabernat/32/22_2.png) [@rabernat](https://discourse.pangeo.io/u/rabernat)\
**Post date:** [November 25, 2024, 1:56pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/4 "2024-11-25T13:56:00Z")

</div>

To help us help you, please share the output of

```python
data = xds.fillna(nodata).astype(dtype)).rio.set_nodata(nodata)
print(data.info())

```

---

<div class="post-metadata">

**Author:** ![remi-braun](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/remi-braun/32/1178_2.png) [@remi-braun](https://discourse.pangeo.io/u/remi-braun)\
**Post date:** [November 25, 2024, 2:13pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/5 "2024-11-25T14:13:52Z")

</div>

Hi @rabernat !

I have a DataArray, not a Dataset, so the `info()` function doesn’t seem to exist 🤔

Here is the print of the DataArray:

```auto
<xarray.DataArray 'COG of EMSR773_AOI17_GRA_CONSOLIDATION_AERIAL_20241111_1300_ORTHO_cog' (
                                                                                           band: 3,
                                                                                           y: 41909,
                                                                                           x: 41362)> Size: 5GB
dask.array<astype, shape=(3, 41909, 41362), dtype=uint8, chunksize=(1, 2048, 2048), chunktype=numpy.ndarray>
Coordinates:
  * band (band) int32 12B 1 2 3
  * x (x) float64 331kB -5.481e+04 -5.481e+04 ... -4.24e+04 -4.24e+04
  * y (y) float64 335kB 4.754e+06 4.754e+06 ... 4.741e+06 4.741e+06
    spatial_ref int32 4B 0
    quantile float64 8B 0.02
Attributes:
    long_name: ['RED', 'GREEN', 'BLUE']

```

---

<div class="post-metadata">

**Author:** ![maxrjones](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/maxrjones/32/3190_2.png) [@maxrjones](https://discourse.pangeo.io/u/maxrjones)\
**Post date:** [November 25, 2024, 4:02pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/6 "2024-11-25T16:02:54Z")

</div>

> This is where I don’t understand, where does this value 999999 comes from?

This is most likely coming from NumPy’s defaults for masked arrays - [numpy.ma.default\_fill\_value — NumPy v2.1 Manual](https://numpy.org/doc/stable/reference/generated/numpy.ma.default_fill_value.html).

Based on the traceback mentioning get\_data and deserialize I think the actual error is probably with reading the data from the original file rather than the writing process. What happens if you call a .load early in the workflow? Do you get the same error?

---

<div class="post-metadata">

**Author:** ![remi-braun](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/remi-braun/32/1178_2.png) [@remi-braun](https://discourse.pangeo.io/u/remi-braun)\
**Post date:** [November 25, 2024, 4:49pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/7 "2024-11-25T16:49:52Z")

</div>

You mean do a `load`, recreate a dask array from the loaded array and resave it with `save_cog_with_dask`?

---

<div class="post-metadata">

**Author:** ![maxrjones](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/maxrjones/32/3190_2.png) [@maxrjones](https://discourse.pangeo.io/u/maxrjones)\
**Post date:** [November 25, 2024, 5:28pm UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/8 "2024-11-25T17:28:43Z")

</div>

My suggestion was a debugging approach rather a final solution.

If there’s no existing shared solution, no one from these forums has encountered the exact same solution (thank you for asking!), and the data cannot be shared, I think you’ll probably need to do some digging to find out what is causing the issue. I was suggesting a mechanism for narrowing down where in the workflow it’s happening, by seeing if you get similar errors with:

```auto
# Test just opening the dataset
xds.load()
# Test filling missing data
xds.fillna(nodata).load()
...

```

Of course you may have already tried this all and only find the issue when passing the output from all the piped operations to `save_cog_with_dask`, so apologies if my suggestion was unhelpful.

---

<div class="post-metadata">

**Author:** ![remi-braun](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/remi-braun/32/1178_2.png) [@remi-braun](https://discourse.pangeo.io/u/remi-braun)\
**Post date:** [November 26, 2024, 7:56am UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/9 "2024-11-26T07:56:07Z")

</div>

What I can say is that is works like a charm when I save this without Dask, with rioxarray (after a compute), with `to_raster(..., driver="COG", windowed=True)`.  
It also seems to work with numpy \< 2.

However, it fails with numpy \>2.  
I tested also with numpy 2.1.0, but the error is slightly different (same topic though)  
`TypeError: Cannot cast scalar from dtype('int64') to dtype('uint8') according to the rule 'same_kind'`

> **Traceback**
>
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\sertit\rasters.py”, line 258, in wrapper  
> raise ex  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\sertit\rasters.py”, line 254, in wrapper  
> out = function(any\_raster\_type, \*args, \*\*kwargs)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\sertit\rasters.py”, line 1181, in write  
> ).compute()  
> ^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\dask\base.py”, line 372, in compute  
> (result,) = compute(self, traverse=False, \*\*kwargs)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\dask\base.py”, line 660, in compute  
> results = schedule(dsk, keys, \*\*kwargs)  
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\dask\array\reductions.py”, line 477, in chunk\_max  
> return np.max(x, axis=axis, keepdims=keepdims)  
> ^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\_core\fromnumeric.py”, line 3199, in max  
> return \_wrapreduction(a, np.maximum, ‘max’, axis, None, out,  
> ^^^^^^^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\_core\fromnumeric.py”, line 84, in \_wrapreduction  
> return reduction(axis=axis, out=out, \*\*passkwargs)  
> ^^^^^^^^^^^  
> File “C:\Users\rbraun\Anaconda3\envs\eoprocesses\Lib\site-packages\numpy\ma\core.py”, line 6091, in max  
> np.copyto(result, result.fill\_value, where=newmask)  
> ^^^^^^^  
> TypeError: Cannot cast scalar from dtype(‘int64’) to dtype(‘uint8’) according to the rule ‘same\_kind’

My idea is that it doesn’t fail with numpy \< 2 because such newly illegal casting was possible before, but not anymore.  
It is therefore more likely an issue and I’ll go to odc-geo creating one.

Thanks for the help 🙏

---

<div class="post-metadata">

**Author:** ![remi-braun](https://yyz2.discourse-cdn.com/flex030/user_avatar/discourse.pangeo.io/remi-braun/32/1178_2.png) [@remi-braun](https://discourse.pangeo.io/u/remi-braun)\
**Post date:** [November 26, 2024, 8:34am UTC](https://discourse.pangeo.io/t/save-cog-with-dask-cannot-convert-fill-value-999999-to-dtype-uint8/4713/10 "2024-11-26T08:34:02Z")

</div>

I created the issue at `odc-geo` and I worked on a minimal working example after all (you can find it there, in order to avoid creating multiple parallel threads about this):  
 → ['save\_cog\_with\_dask' fails with numpy \> 2 · Issue #189 · opendatacube/odc-geo · GitHub](https://github.com/opendatacube/odc-geo/issues/189)