Version 0.2.1

What's Changed

• Added a masking token for pruned vocabulary by @cjweir in #47
• added token frequency to the kernel functions by @cjweir in #48
• Skip n-grams and better handling of directional windowing (including support for non-symmetric windows) via "reverse". by @lmcinnes in #50
• Kernel and Window args by @cjweir in #49
• Kernel offsets by @cjweir in #51
• Linear Optimal Transport by @lmcinnes in #52
• EMTokenCooccurrence update by @cjweir in #53
• EMTCV skip_grams by @cjweir in #54
• Fixes for spherical vectors (e.g. word vectors) by @lmcinnes in #55
• Add simple transformers by @lmcinnes in #56
• Refactoring by @cjweir in #57
• ApproxWassersteinVectorizer (LA vectorizer), supervised InformationWeight by @lmcinnes in #58
• BayesEM and Dynamic COO memory by @cjweir in #59
• Handle disabling JIT for to_fixed_tuple by @lmcinnes in #60
• added document_context to TokenCooccurrenceVectorizer() by @jc-healy in #61
• Change default normalization in TCV by @lmcinnes in #63
• Document_context EM by @cjweir in #65
• Added a summarize_embedding function to vectorizers.utils. by @jc-healy in #66
• Refactored the variable document_context to the more intuitively name… by @jc-healy in #67
• added a helper function to utils called categorical_columns_to_list. by @jc-healy in #68
• Add SinkhornVectorizer for larger distributions by @lmcinnes in #69
• CategoricalColumnTransformer documentation by @jc-healy in #70
• Add dask for simple parallel computation in the word vectorizer by @lmcinnes in #71
• Count feature compressor by @lmcinnes in #72
• Fix doc builds per new docs required packages by @lmcinnes in #73
• Pandoc now required for doc build by @lmcinnes in #74
• Make some changes to how info weight works; New README. by @lmcinnes in #75
• SlidingWindowTransformer for working with time-series like data by @lmcinnes in #76
• Documentation Skeleton by @lmcinnes in #77
• Add document_vectorization basic tutorial by @lmcinnes in #78
• More documentation updates by @lmcinnes in #79
• Added a sep parameter to utils.summarize_embedding() to allow for dif… by @jc-healy in #80
• Generator based WassersteinVectorizer for sliding window text embeddings by @lmcinnes in #81
• Fix up some of the test coverage. by @lmcinnes in #82
• Refactor transformers by @lmcinnes in #83
• SequentialDifferenceTransformer, function kernels by @lmcinnes in #85
• Add named function kernels for sliding windows by @lmcinnes in #86
• Add Compression Vectorizers by @lmcinnes in #87
• added EdgeListVectorizer by @jc-healy in #88
• Added max_unique_tokens by @cjweir in #89
• Ensure contiguous arrays in optimal transport vectorizers by @lmcinnes in #90
• Ignore Jupyter artifacts by @hamelin in #92
• Add unit test for CFC transformer by @hamelin in #91
• Fix documentation by @cakiki in #95
• Fixed categorical column transformer example in our documentation by @jc-healy in #98
• [Typo] Rename enviornment.yml to environment.yml by @cakiki in #100
• The big cooccurrence refactor by @cjweir in #99
• TimedTCV bug fix by @cjweir in #102
• Some documentation cleanup by @lmcinnes in #103
• [Typo] Update README.rst by @cakiki in #104
• Added some taxonomy of sequences documentation. by @jc-healy in #105
• Testing how well an html page integrates with our sphinx documentation. by @jc-healy in #106
• Revert "Testing how well an html page integrates with our sphinx documentation." by @jc-healy in #107
• Added SignatureVectorizer (iisignature) by @jh83775 in #108
• Overloaded addition for NgramVectorizer for easy batch processing by @jc-healy in #110
• Take iisignature off of the list of direct deps by @hamelin in #112
• Bump version to 0.2 by @hamelin in #113
• Update preprocessing.py by @cakiki in #116
• [MINOR] Update README.rst by @cakiki in #117
• Minor fixes by @lmcinnes in #120
• Fixed a memory allocation bug in MultiSetCooccurrenceVectorizer adde… by @jc-healy in #119
• Add publishing step to devops config by @gclendenning in #121
• Add additional branch triggers by @gclendenning in #125
• Ensure all tests pass before release by @gclendenning in #127
• Refactored WassersteinVectorizer by @jc-healy in #126
• updated version number to push to pypi by @jc-healy in #129
• Fix version number by @hamelin in #131

New Contributors

• @hamelin made their first contribution in #92
• @cakiki made their first contribution in #95
• @jh83775 made their first contribution in #108
• @gclendenning made their first contribution in #121

Full Changelog: v0.01...release-0.2.1