Closed GenevieveBuckley closed 1 year ago
Here's a recent CI workflow from the main branch (I triggered this manually from the Actions tab). We see the same problem with the versions in the CI. https://github.com/dask/dask-image/actions/runs/4738096466/jobs/8411587073
Noting that this should be resolved with https://github.com/conda-forge/conda-forge-repodata-patches-feedstock/pull/451, which makes it so Dask packages prior to 2023.2.0 now have a pandas<2
constraint
Thank you @charlesbluca
There is an incompatibility between pandas 2.0.0 and dask 2022.1.1, causing our CI to break.
It's recommended to upgrade Dask to >=2023.2.0 in order to use pandas >=2.0.0 https://github.com/dask/dask/issues/10164
To avoid this problem affecting dask-image, we should pin/upgrade/etc...
Here's a copy-paste of the relevant section of the CI log
Details (click to expand)
``` 2023-04-17T03:04:25.7343567Z Running setup.py develop for dask-image 2023-04-17T03:04:26.1248592Z Successfully installed dask-image-0+untagged.1.g5b16ff7 2023-04-17T03:04:27.2898310Z # packages in environment at /usr/share/miniconda3/envs/dask-image-testenv: 2023-04-17T03:04:27.2899254Z # 2023-04-17T03:04:27.2900491Z # Name Version Build Channel 2023-04-17T03:04:27.2901242Z _libgcc_mutex 0.1 conda_forge conda-forge 2023-04-17T03:04:27.2901820Z _openmp_mutex 4.5 2_gnu conda-forge 2023-04-17T03:04:27.2902501Z aom 3.5.0 h27087fc_0 conda-forge 2023-04-17T03:04:27.2903771Z attrs 22.2.0 pyh71513ae_0 conda-forge 2023-04-17T03:04:27.2904195Z blosc 1.21.3 hafa529b_0 conda-forge 2023-04-17T03:04:27.2904657Z bokeh 2.4.3 pyhd8ed1ab_3 conda-forge 2023-04-17T03:04:27.2905035Z brotli 1.0.9 h166bdaf_8 conda-forge 2023-04-17T03:04:27.2905401Z brotli-bin 1.0.9 h166bdaf_8 conda-forge 2023-04-17T03:04:27.2905770Z brunsli 0.1 h9c3ff4c_0 conda-forge 2023-04-17T03:04:27.2906127Z bzip2 1.0.8 h7f98852_4 conda-forge 2023-04-17T03:04:27.2906691Z c-ares 1.18.1 h7f98852_0 conda-forge 2023-04-17T03:04:27.2907044Z c-blosc2 2.8.0 hf91038e_1 conda-forge 2023-04-17T03:04:27.2907419Z ca-certificates 2022.12.7 ha878542_0 conda-forge 2023-04-17T03:04:27.2907797Z certifi 2022.12.7 pyhd8ed1ab_0 conda-forge 2023-04-17T03:04:27.2908156Z cfitsio 4.2.0 hd9d235c_0 conda-forge 2023-04-17T03:04:27.2908513Z charls 2.4.1 hcb278e6_0 conda-forge 2023-04-17T03:04:27.2908941Z click 8.0.4 py39hf3d152e_0 conda-forge 2023-04-17T03:04:27.2909367Z cloudpickle 2.2.1 pyhd8ed1ab_0 conda-forge 2023-04-17T03:04:27.2909827Z contourpy 1.0.7 py39h4b4f3f3_0 conda-forge 2023-04-17T03:04:27.2910275Z coverage 6.3 py39h3811e60_0 conda-forge 2023-04-17T03:04:27.2910671Z cycler 0.11.0 pyhd8ed1ab_0 conda-forge 2023-04-17T03:04:27.2911268Z cytoolz 0.12.0 py39hb9d737c_1 conda-forge 2023-04-17T03:04:27.2911726Z dask 2022.1.1 pyhd8ed1ab_0 conda-forge 2023-04-17T03:04:27.2912174Z dask-core 2022.1.1 pyhd8ed1ab_0 conda-forge 2023-04-17T03:04:27.2912871Z dask-image 0+untagged.1.g5b16ff7 dev_0Originally posted by @GenevieveBuckley in https://github.com/dask/dask-image/issues/311#issuecomment-1513946913