v2.6.1

Fix

• Updated dev. dependencies to not include upper bounds (48bedec)

Documentation

• Removed error from notebook (708d9d7)
• Added description of readability scores (16cda18)

v2.6.0

Feature

• Add sklearn transformer (be733c6)

Documentation

• Add sklearn to requirements in docs (33194c9)
• Minor docstring changes (375d3e0)
• Add sklearn tutorial (4220072)

v2.5.1

Fix

• Don't subtract 1 from counts if add_all_tags is False (02b61a6)

v2.5.0

Feature

• Add proportions of all pos tags (664c0e8)

Fix

• Listed metrics deviate between extraction functions in docs (4632407)
• Handle empty strings in pos_proportions (762b7b2)

v2.4.6

Fix

• Bump version (b59cdd1)

v2.4.5

Fix

• Dep dist. test (a3b54a6)
• Coherence model now handles empty strings as intended (abfa507)
• Dep distance test (2bed2c1)

Documentation

• Added demo to docs (6121eeb)

v2.4.4

Fix

• Add .zenodo.json (3cc463c)

Documentation

• Add test requirements to faq (1550de7)

v2.4.3

Fix

• Add descriptive_stats pipe no matter the verbosity (3fd3240)
• Change default verbosity of readability component (ddfd269)
• Change verbosity of lexeme_prob_table (b64631f)
• Automatically download supplied spacy model if not on disk (f54a9b6)

Documentation

• Change quick start model to lg (653fb49)
• Update readme to use lg model for proper coherence calculation (cdd85b7)

v2.4.2

Fix

• Handle case where extension is set, but pipeline is not (2ba2dd2)
• extract_metrics failing if metrics=None (6bfe6b0)

v2.4.1

Fix

• Change auto-approve event (729836f)