--multir

Switch

--multir

Description

Multivariate regression: regresses each feature on all requested outcomes simultaneously (one model per feature, all outcomes as predictors — the reverse direction of --correlate, which fits one model per outcome/feature pair).

Argument and Default Value

No argument. Boolean switch, default: False.

Details

Calls oa.multRegressionWithFeatures(fg, spearman, p_correction_method, blacklist, whitelist, showfeatfreqs, outcomeWithOutcome, interactions=args.interactions). This is the switch --interactions requires — without --multir, --interactions has no effect. With --csv, writes results via oa.writeSignificantCoeffs4dVis() to a ..multiprint.. file for use with 4dVis-style visualization, using --output_name/--output_dir the same way as other correlate-family output.

Other Switches

Required Switches:

• -d, -g, -t

• -f

• --outcome_table, --outcomes

Optional Switches:

• --interactions

• --csv

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 gender --multir --interactions --csv --output_name ./multir_out