Command Line Execution - DeepPhe/DeepPhe-Release GitHub Wiki

A DeepPhe pipeline can be run using the Piper File Submitter GUI that comes with the DeepPhe distribution, it can also be run using a command line in a standard terminal.

Simply open a terminal, navigate to your DeepPhe installation's .DeepPhe/ directory where the DeepPhe NLP application is installed, and execute the runDeepPhe script using:

./bin/runDeepPhe.sh -i <InputDirectory> -o <OutputDirectory>

The command for windows is nearly identical:

bin/runDeepPhe.bat -i <InputDirectory> -o <OutputDirectory>

Input Directory

The input directory should contain a corpus with one directory per patient, with that patient's text documents inside the patient directory.

example_corpus/
  Patient_001/
    note_001.txt
    note_002.txt
  Patient_002/
    note_001.txt

Output Directory

The output directory is a root directory that should contain files produced by your pipeline. Not that files may not all be written in the output directory; some files may be written in subdirectories of the output directory based upon other settings for your pipeline.