v1.107.2

New features

• #302 Added support for show_available_date option.

Internal changes

• #296 Improve testing of our sample Jupyter notebooks.
• #299 Fix links in at_time demo notebook.

v1.101.0

Improvements

• added include_names option to GroClient.get_df to include human-readable names in output #236
• renamed unicode filename, hopefully fixes issue installing our Conda package on Windows (d97c5c1)
• added __version__ to module #298

v1.100.0

Improvements

• full rewrite of similar regions example code #227
• fix bug that caused fetch() called on closed AsyncHTTPClient errors when creating multiple GroClients #297
• retry on 502 responses #285

Housekeeping/internal improvements:

• documentation fixes #284, sphinx update #292, docs migrated docs checked by CI #294

v1.98.0

New Features

• Automated PyPI releases (#277, #281) . This package is now installable via pip install groclient

Improvements

• Added deprecation warnings for api modules (#278, #283).
• Converting conf_interval value into requested unit (#279), and make it to snake_case (#280).

v1.83

New features

• general get_descendant() convenience function to traverse entity graph for items and metrics (in addition to previously supported get_descendant_regions())

Improvements

• move GroClient from api/client to top level groclient module (backward compatible)
• avoid duplicates in rank_series_by_source

v1.82

New Features

• Adding support for zoom levels in get_geojsons()

Improvements

• Simplified client class options (BatchClient merged into GroClient)

v1.77.1

• get_geojsons fix.

v1.77.0

New Features

GroClient.get_geojsons enables retrieval of geometry information for region trees, rather than just single regions.

v1.76.0

New Features

• GroClient.lookup() can now accept lists of ids in a single request. This can speed up retrieving details about many entities at once.
• New GroClient.get_top() method allows you to get the items or regions with the greatest values for the given selections. Great for questions like "what countries produce the most corn?" or "what are India's major exports?"

Improvements

• Now, you can use df.index.names with GroClient.get_df() to identify the indices. For example:

  >>> client.add_single_data_series({'metric_id': 8671080, 'item_id': 274, 'region_id': 1215, 'partner_region_id': 0, 'source_id': 63, 'frequency_id': 9})
  >>> df = client.get_df()
  >>> dict(zip(df.index.names, df.iloc[0].name))
  {'metric_id': 8671080, 'item_id': 274, 'region_id': 1215, 'partner_region_id': 0, 'frequency_id': 9, 'source_id': 63}

• Performance improvements to GroClient.get_descendant_regions(), GroClient.search_and_lookup(), and GroClient.lookup_belongs()

v1.71.0

New Features

• optional result_filter in GroClient.find_data_series() gives users deeper control of ambiguous search results
• optional weighting_func in CropModel.compute_crop_weighted_series() allows user to specify custom weighting function
• index_by_series option in GroClient.get_df() gives dataframe access indexed by data series.

Improvements

• improved search in GroClient.find_data_series() (and CropModel.compute_crop_weighted) , now using frequency and source ranking by coverage
• optimized data retrieval in GroClient.get_df() for repeated use