6 Pull requests merged by 3 people

• Update Dask Cloud Provider documentation to include Nebius as a supported cloud option

  #11703 merged Jan 28, 2025

• Fix normalize_chunks when squashing into a single chunk

  #11702 merged Jan 28, 2025

• Fix positional indexing with newaxis

  #11699 merged Jan 27, 2025

• Set array backend in scipy-sparse-indexing

  #11700 merged Jan 24, 2025

• Fix value_counts shuffling strategy

  #11698 merged Jan 24, 2025

• Disentangle core expression class from dataframe specific code

  #11688 merged Jan 22, 2025

3 Pull requests opened by 2 people

• Implement infrastructure, random, blockwise and Elemwise

  #11689 opened Jan 22, 2025

• Implement unify_chunks and Rechunk

  #11692 opened Jan 23, 2025

• array: Never share buffers after compute()

  #11697 opened Jan 24, 2025

6 Issues closed by 4 people

• `bind` parameter for preserving keys of the regenerated nodes

  #11701 closed Jan 27, 2025

• Array slicing with boolean mask and `np.newaxis` stopped working at some point

  #11694 closed Jan 27, 2025

• Removal of legacy DataFrame broke ability to construct DataFrame from dask tasks

  #11678 closed Jan 24, 2025

• `cupyx.scipy.sparse`: Unsupported initializer format

  #11696 closed Jan 24, 2025

• `.value_counts()` failed to aggregate across partitions

  #11695 closed Jan 24, 2025

• Passing keyword arguments to delayed async functions and computing it

  #11690 closed Jan 24, 2025

2 Issues opened by 2 people

• Performance regression from v2024.7.1 to 2024.8.0 when resampling in xarray

  #11693 opened Jan 23, 2025

• Errors with Zarr v3 and da.to_zarr()

  #11691 opened Jan 23, 2025

14 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• `map_blocks` shape and chunk calculations are wrong when combining chunk aligned but different `ndim` arrays

  #11686 commented on Jan 22, 2025 • 0 new comments

• Unable to use scheduler 'external_address'

  #11681 commented on Jan 22, 2025 • 0 new comments

• dask.delayed: Dataframe partition with map_overlap gets randomly converted to Pandas Dataframe or Pandas Series.

  #11200 commented on Jan 22, 2025 • 0 new comments

• Regarding additional feature on Incremental SVD over available SVD_COMPRESSED method.

  #11527 commented on Jan 25, 2025 • 0 new comments

• Switch from numpy.fft to scipy.fft?

  #6440 commented on Jan 27, 2025 • 0 new comments

• Documentation for custom Aggregation does not cover NA edge case

  #6410 commented on Jan 27, 2025 • 0 new comments

• HDFS remote data source refers to deprecated API

  #6400 commented on Jan 27, 2025 • 0 new comments

• Use parquet metadata to get length

  #6387 commented on Jan 27, 2025 • 0 new comments

• When combining complex column selections on dataframes, join the selection and push down

  #6388 commented on Jan 27, 2025 • 0 new comments

• Sort dask array

  #4368 commented on Jan 28, 2025 • 0 new comments

• Allow the deferral of array.to_hdf5

  #6405 commented on Jan 27, 2025 • 0 new comments

• Use automatic chunking in from_zarr

  #6419 commented on Jan 27, 2025 • 0 new comments

• Convert dataframes to bags with map_partitions

  #6448 commented on Jan 27, 2025 • 0 new comments

• Reduce pickle size for tasks

  #11687 commented on Jan 28, 2025 • 0 new comments