mtlynch / ingredient-phrase-tagger

57%

master: 72%

LAST BUILD ON BRANCH ignore-index

branch: ignore-index

CHANGE BRANCH

x

Reset

• ignore-index
• 20180503_173729
• coverage-demo
• coverage-first
• coverage-poc
• cut-except
• cut-pandas
• data-reader
• data-writer
• default-model
• do-the-shuffle
• docstring
• easy-parse
• fix-coverage
• fix-label-order
• fix-train
• fractional-split
• handle-empty
• handle-ml
• link-to-zestful
• master
• more-coverage
• more-e2e
• more-tokenizing
• partitioner2
• readme-links
• remove-bootstrap
• remove-cruft
• test-travis
• testing-output
• tokenize
• train-prod-model
• train-script
• train-whitespace
• travis-key
• unicodecsv
• unidecode
• unit-tests
• upgrade-pyflakes
• writerows
• xplatform-readme

Build # 96

Build Type

push

travis-ci

Committed by mtlynch

Commit Message

Ignoring 'index' field for tags

This fixes a bug where the tagger would sometimes tag tokens with B-INDEX based on the 'index' column of the CSV file, but that column is unrelated to the data itself (just a count of rows) and should never be used as a tag.

Run Details

17 of 17 new or added lines in 1 file covered. (100.0%)

110 of 193 relevant lines covered (56.99%)

0.57 hits per line