github Kotlin/dataframe 1.0.0-rc01

latest releases: build-10866, build-10865
4 hours ago

Kotlin DataFrame 1.0.0-rc01 is released!

What's new?

  • All our example projects now use the compiler plugin.
  • requireColumn - addition to cast, convertTo family of functions. See requireColumn documentation for detailed example.
  • Documentation pages for pivot and groupBy provide code examples for all possible aggregations - 170 in total and all with rich visualization.
  • Updated quickstart guide shows a few simple steps: loading data, transforming it, and visualizing it in IntelliJ IDEA project.
  • Changes in DataFrame.print to make it more comprehensive: rendering column types and total col x row by default; its rowsLimit and valuesLimit parameters now accept null value to disable truncation.

And many other improvements and fixes! See below for a complete list of changes in this release grouped by category.

We invite you to try this release:

  • Use the 1.0.0-rc01 version in your Gradle or Maven projects:
  • Use the special 1.0.0-rc01n version in Kotlin Notebook. The n version includes a patch that resolves notebook specific issue with statistical functions.
    • You can specify the version in the descriptor or use %useLatestDescriptors.
      If you want to include kandy, list it after dataframe:
      %useLatestDescriptors
      %use dataframe, kandy
      // or
      %use dataframe(1.0.0-rc01n), kandy(0.8.5, 0.8.5)
    • Use kernel version 0.16.0-736 or newer.

Features

  • Added requireColumn operation by @koperagen in #1715
  • Added DataFrame.renderToMarkdown(): String function by @koperagen in #1760
  • Made emptyDataFrame reified so resulting "null object" dataframes are suitable for aggregation and other column access operations like select by @koperagen in #1914
  • Added countDistinct overloads on GroupBy by @Allex-Nik in #1875
  • Added Json instance overload to readJson(inputStream, ...) by @Jolanrensen in #1869
  • Improved rendering in DataFrame.print() by @koperagen in #1760
    rowsLimit , valuesLimit parameters now accept null to display all rows or complete column values.
    New borderStyle parameter to choose what ASCII symbols will be used for border
    Types are now rendered on a separate row, and rendering is enabled by default
  • Aligned column name clashing handling across different IO by @AndreiKingsley in #1904
  • Supported reading new GEOMETRY, TIME_NS, VARIANT types from DuckDB by @Jolanrensen in #1924

Removals and Deprecations

  • Removed remaining data modifiers from intermediate operations classes: Corr, SplitWithTransform for cleaner public API by @koperagen in #1826
  • rename into is deprecated in favor of rename to by @Jolanrensen in #1837
  • Deprecated format guessing DataFrame.read() and SupportedDataFrameFormat in favor of specific DataFrame.readCsv/readJson/... functions by @Jolanrensen in #1916

Compiler plugin

  • Annotated DataColumn.map* for compiler plugin support by @koperagen in #1825
  • Annotated DataFrame.get(String columns) for compiler plugin support by @koperagen in #1853
  • Annotated allExcept CS DSL operations for compiler plugin support by @Jolanrensen in #1897
  • Updated DataFrame.concat signature for better plugin support
  • Annotated DataFrameBuilder APIs for plugin support by @koperagen in #1828

Fixes and other changes

Documentation and Examples

Full Changelog: v1.0.0-Beta5...1.0.0-rc01

Don't miss a new dataframe release

NewReleases is sending notifications on new releases.