Major changes in this release: * New interactive Jupyter notebook widget for visualizing and modifying DataFrames containing one or more columns of span data. See demo notebook [here](https://github.com/CODAIT/text-extensions-for-pandas/blob/master/notebooks/DataFrame_Widget_Demo.ipynb). * Improvements to Jupyter notebook visualizations of `SpanArray` objects * Enhanced support for reading CoNLL-U files. Semantic role labeling data from [Universal Propositions](https://github.com/System-T/UniversalPropositions) now supported. * Improved [API documentation](https://text-extensions-for-pandas.readthedocs.io/en/latest/)
Minor changes: * Updated workaround for https://github.com/pandas-dev/pandas/issues/42430 (see https://github.com/CODAIT/text-extensions-for-pandas/pull/236) * Updated development environment build process to maintain compatibility with Python 3.6. Note that some of our demo notebooks no longer work on that version of Python because other libraries they depend on have dropped Python 3.6 support.
0.1.1
Minor release that fixes issue 231. This release is otherwise identical to version 0.1.
0.1
Major changes in this release: * First non-beta release * Support for Pandas 1.3.0 (217, 219, 218) * Improved visualization of SpanArrays in Jupyter (201, 202) * Improved support for common variants of CoNLL-U format (191, 193, 195, 194, 198)
0.1b4
Beta release with minor updates: * Add Python 3.6 to the range of supported Python versions (192) * Add NumPy ufunc to TensorArray (187)
0.1b3
Beta release. Includes the following updates: * Support for SpanArrays and TokenSpanArrays that cover multiple documents (73) * Feather I/O for multi-document SpanArrays (179) * Improved support for Pandas 1.2.x (159, 159) * Changed HTML rendering so that spans will render correctly in JupyterLab dark mode and in VSCode notebooks (part of 170) * Added an API for online creation of dictionaries and a SpaCy tokenizer configuration for dictionary matching (180)