--categories

Switch

--categories CATEGORY [CATEGORY ...]

Description

Restricts a lexicon-driven black/white list (see --blacklist/--whitelist) to specific categories from the lexicon table given by -l.

Argument and Default Value

One or more category names/ids as they appear in the lexicon table's category column. Default: none (all categories used).

Details

Verified: used together with -l and --whitelist (or --blacklist), this builds a white/black list of the individual words belonging to the given lexicon categories, for filtering a word/ngram-level feature table (e.g. feat$1gram$...) — not for filtering an already-built category-level table (feat$cat_...), which has no use for this switch since its features already are the category names. Pass * to mean all categories in the lexicon table.

Other Switches

Required Switches:

-l LEX_TABLE
--whitelist or --blacklist

Example Commands

dlatkInterface.py -d dla_tutorial -t msgs -c user_id -f 'feat$1gram$msgs$user_id$16to16' \
--outcome_table blog_outcomes --outcomes age -l met_a30_2000_cp --categories 344 --whitelist --correlate