--p_value
Switch
--p_value
Description
P value to use when creating tagclouds and topic tagclouds and for printing number of significant features to terminal when running correlate.
Argument and Default Value
Also used with mediation analysis: specify the significance level of results to be printed in the summary report. Any mediation results with Sobel p less than the specified value will be printed.
Details
- When using --correlate this switch will print the number of features significant at this threshold:
#Example output 1259 features significant at p < 1e:doc:fwflag_05
Other Switches
Optional Switches: --correlate --tagcloud --topic_tagcloud --DDLATagcloud? Example Commands ================
Using the packaged dla_tutorial dataset (see Differential Language Analysis (DLA) Tutorial):
# Create topic tagclouds
dlatkInterface.py -d dla_tutorial -t msgs -c user_id --group_freq_thresh 500 \
-f 'feat$cat_met_a30_2000_cp_w$msgs$user_id$16to16' --outcome_table blog_outcomes --outcomes age \
--p_correction simes --topic_tagcloud --make_topic_wordclouds \
--topic_lexicon met_a30_2000_freq_t50ll --csv --output_name ~/xxx_output --p_value 0.001
# Correlate features with outcome
dlatkInterface.py -d dla_tutorial -t msgs -c user_id --group_freq_thresh 500 \
-f 'feat$cat_met_a30_2000_cp_w$msgs$user_id$16to16' --outcome_table blog_outcomes --outcomes age \
--no_correction --p_correction simes --correlate --p_value 0.001