--load_model

Switch

--load_model

Description

Loads the model that is stored in a file.

Argument and Default Value

None

Details

If you are running a command that uses a model that was trained previously, using this will tell the infrastructure to load that model. This switch requires the use of --picklefile. In order to use this, you must have saved the model using --save_model. See one of these three for more info: --train_regression --train_classifiers --cca

Other Switches

Required Switches:

• -d, -g, -t,

• -f,

• --outcome_table, --outcomes

• --picklefile

Optional Switches: * --predict_regression and anything optional to that * --predict_classifiers and anything optional to that * --cca and anything optional to that etc.

Example Commands

Using the packaged dla_tutorial dataset (see Applying A Pickle Model): loads a previously trained regression model and uses it to predict age for a new set of users.

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' 'feat$1to3gram$msgs$user_id$16to16$0_1' 'feat$1to3gram$msgs$user_id$16to1$0_1' \
--outcome_table ageDummy --outcomes age --predict_regression_to_feats lbp_age \
--load_model --picklefile ~/age.2000fbtopics.1to3grams.16to16.16to1.ridge1000.magic_sauce.gft500.pickle