classtf

classtf is a Random Forest Classifier based on Tensorflow (TM)

Usage

classtf.py [-h] [-r FILE] [-t FILE] [-x FILE] [-f] [--loss-treshold VALUE] [--train-timeout TIME_INTERVAL] [-c TARGET_FEATURE_ID] [--ignore-features [...]] [-d MODEL_DIR] [-n NUM_OF_TREES] [-b BATCH_SIZE] [--depth NUM_NODES] [-v]

Arguments

-h, --help

show this help message and exit

-r FILE, --run FILE

If specified, the classifier is run using the dataset FILE as input

-t FILE, --train FILE

If specified, the classifier is trained using the dataset FILE

-x FILE, --test FILE

If specified, the classifier is tested using the dataset FILE

-f, --feature-importance

If specified, the importance of each feature is computed. Can be used only if both --train and --test options are specified

--loss-treshold VALUE

If specified, the training will stop when the loss changes between two cycles becomes smaller than VALUE (or when the training timeout expires). This option has effect only when the option --train is specified and is ignored otherwise. If not specified, the default value of 0.001 is used.

--train-timeout TIME_INTERVAL

If specified set the maximum execution time for the training process. TIME_INTERVAL must be a string representing a time interval. Allowed units are y (years), d (days), h (hours), m (minutes), s (seconds) [i.e. 1y2d13h20m13.3s]

If not specified, no timeout is applied

-c TARGET_FEATURE_ID, --target TARGET_FEATURE_ID

Set the name or column index of feature used as target class during training and testing. If not specified, the last column in the dataset is used as default

--ignore-features [ ...]

List of features that should be ignored

-d MODEL_DIR, --model-dir MODEL_DIR

If specified, the trained model is saved or restored from MODEL_DIR

-n NUM_OF_TREES, --trees NUM_OF_TREES

If specified, set the number of generated trees to NUM_OF_TREES, otherwisee fallback to the default value of 1000 trees

-b BATCH_SIZE, --batch-size BATCH_SIZE

If specified, set the size of the batch to to BATCH_SIZE, which is the number of object used at once during a training/test/run cycle. The default value is 4096

--depth NUM_NODES

If specified, set the maximum number of nodes created by the model to NUM_NODES. The default value is 10000

-v, --version

Print the program version and exit

NOTES:

You MUST specify at least one of the -r, -t or -x options

Examples

Simple training and testing: classtf.py -t cat1.fits -x cat2.fitsi -c "Class"
Train the classifier and save the trained model: classtf.py --train traincat.fits --model-dir ./mymodel/
Load saved model and run it on a catalogl: classtf.py --model-dir ./mymodel --run mycat.fits
Train the classifier with a timeout of 1 day and 6 hours classtf.py --train traincat.votable --train_timeout 1d3h classtf.py --train traincat.votable --train_timeout 1.25d classtf.py --train traincat.votable --train_timeout 30h
Train the classifier with a loss treshold of 0.001: classtf.py --train train.csv --loss-treshold 0.001

For more info import the classtf module in python and run

 >>> help(classtf)

mauritiusdadd/classtf