PRIMAP2 Version 0.11.2

primap2 0.11.2 (2024-10-07)

primap2 is a library for compiling and analyzing climate policy datasets.

Changes

Improvements

• Added support for python 3.12.

  Started testing the lowest supported versions of dependencies. (#259)

• Added support for and required pandas 2.2.

  Switched accelerated filling functions from bottleneck to numbagg. (#261)

Bug Fixes

• Fixed handling of NaN values in {py.func}xarray.DataArray.pr.set() when existing=overwrite. (#225)

Improved Documentation

• Enhanced the API documentation. (#264)

PRIMAP2 Version 0.11.1

0.11.1

• Workaround for xarray's additional coordinate bug
• Improved a few error messages
• Disable xdoc as it throws errors
• Fixed reading the interchange format with dimensionless data (where the unit is
  an empty string).

PRIMAP2 Version 0.11.0

0.11.0

• Removed Python 3.9 support.
• Added exclude_result and exclude_input parameters to priority definitions for
  compose. They can be used to skip processing entire result timeseries or specific
  input timeseries, e.g. because of invalid data or categories.
• Added protocol for skipping strategies when they aren't applicable for
  input timeseries using the StrategyUnableToProcess exception.
• Added support for specifying the entity in priority and strategy selectors.
• Added support for specifying multiple values in priority and strategy selectors.
• Added negative selection using the primap2.Not value when using the loc
  accessor.
• Use ruff formatting instead of black formatting.
• Added a new csg sub-module which contains functions and configuration models to
  generate a composite dataset from multiple input datasets.
• add global least squares strategy
• add function to add dimension to DatasetDataFormatAccessor
• add functions to aggregate values for coordinates / dimension and entities
  to aggregation accessors
• Added metadata variables to the primap2 data format. Metadata variables describe
  the processing steps done to derive the data on a timeseries level. Also added the
  metadata classes used for the description to the public API. We support saving
  datasets with metadata variables to netcdf, but converting to the interchange format
  looses the metadata variables.
• Explicitly required supported python versions.

PRIMAP2 Version 0.10.0

0.10.0

• Fixed compatibility with latest pint and pint-xarray libraries.

PRIMAP2 Version 0.9.8

0.9.8

• add additional control over nan treatment to downscaling functionality
• Allow kwargs in gas basket summation
• use min_count=1 as default in pr.sum
• fix error message for 0-dimensional arrays in pr.merge
• fix building the documentation on readthedocs.org
• Modify unit harmonization to return native units if possible

PRIMAP2 Version 0.9.7

0.9.7

• Fix the test suite to work with Pint release 0.21.

PRIMAP2 Version 0.9.6

0.9.6

• Add dependency on openscm_units > 0.5.1 for compatibility with latest pandas.
• stop building pdf output documentation, it doesn't support SVG and isn't used much.
• Drop support for Python version 3.8 to prepare for it being dropped
  in Numpy on April 14.

PRIMAP2 Version 0.9.5

0.9.5

• pr.merge: use xarray's combine_first instead of own algorithm for better performance.
• Fix in nan handling of to_interchange_format
• made regexp_unit optional in nir_add_unit_information as intended already before
• Add support for Python 3.11

PRIMAP2 Version 0.9.4

0.9.4

• Update to work with Pint 0.20.1.

PRIMAP2 Version 0.9.3

0.9.3

• Refactor pr.merge.
• Fix primap2 to work with xarray version 2022.06.
• Pin Pint to version 0.20 to work around hgrecco/pint#1631 for now.