List view

• Speed-of-light IO in libcudf

  No due date

  •5/8 issues closed
  62% complete3 open5 closed

• cudf.pandas API coverage

  No due date

  •37/55 issues closed
  67% complete18 open37 closed

• Proxying - cudf.pandas

  No due date

  •24/33 issues closed
  72% complete9 open24 closed

• Profiler - cudf.pandas

  Tasks related to improving cudf.pandas profiler.

  No due date

  •0/3 issues closed
  0% complete3 open0 closed

• Language model acceleration

  No due date

  •4/6 issues closed
  66% complete2 open4 closed

• Benchmarking

  No due date

  •13/24 issues closed
  54% complete11 open13 closed

• Aggregations continuous improvement

  No due date

  •3/16 issues closed
  18% complete13 open3 closed

• Joins continuous improvement

  No due date

  •5/15 issues closed
  33% complete10 open5 closed

• Helps developer velocity

  No due date

  •6/12 issues closed
  50% complete6 open6 closed

• Helps libcudf C++ integrations

  No due date

  •11/25 issues closed
  44% complete14 open11 closed

• ORC continuous improvement

  No due date

  •23/39 issues closed
  58% complete16 open23 closed

• Parquet continuous improvement

  No due date

  •72/115 issues closed
  62% complete43 open72 closed

• Expression evaluation

  No due date

  •24/35 issues closed
  68% complete11 open24 closed

• Stabilizing large workflows (OOM, spilling, partitioning)

  No due date

  •12/20 issues closed
  60% complete8 open12 closed

• Enable streams

  Enable safe usage of a stream-ordered libcudf API

  No due date

  •40/42 issues closed
  95% complete2 open40 closed

• Refactor using cuco containers

  No due date

  •9/12 issues closed
  75% complete3 open9 closed

• Nested JSON reader

  Data-parallel reader for nested JSON text data

  No due date

  •85/107 issues closed
  79% complete22 open85 closed

• CSV continuous improvement

  No due date

  •6/37 issues closed
  16% complete31 open6 closed

• UDF Enhancements

  No due date

  •6/10 issues closed
  60% complete4 open6 closed

• cuDF Python Refactoring

  Refactor Python layers to improve performance, robustness, and extensibility. Ongoing development is being guided by the roadmap and evolving development guide.

  No due date

  •109/115 issues closed
  94% complete6 open109 closed

• Pandas API Alignment and Coverage

  No due date

  •152/243 issues closed
  62% complete91 open152 closed

• List and Struct data types and operations

  No due date

  •53/68 issues closed
  77% complete15 open53 closed

• Time Series Analysis

  No due date

  •38/43 issues closed
  88% complete5 open38 closed