MaxMcGlinnPoole / TermDocumentTensor

For creating a Term Document Tensor
1 stars 2 forks source link

Add command-line functionality #9

Closed donnell794 closed 6 years ago

donnell794 commented 7 years ago

Any ideas on what command-line options and flags we should implement?

MaxMcGlinnPoole commented 7 years ago

some ideas:

-d "name of directory to read files from" -t "read text data" -b "read binary data" -v "visualize the data, should take in name of the visualization" -parafac "use PARAFAC decomposition" -tucker "use Tucker3 decomposition" -o "produce csv output, this will probably need to be more complicated"

I would recommend looking into argparse for doing this: https://docs.python.org/3/howto/argparse.html

MaxMcGlinnPoole commented 7 years ago

Sumanth, what is the update on this?

sumanthneerumalla commented 6 years ago

I've added the options, the variable names for these options can be found in the main function where I did the arg parsing.

Sample usage now: python3 vx.py -d myDirectory -v heatmap -b -parafac -o

Where we specify to parse files in "myDirectory" , output a heatmap, while parsing the files as binary files and using a parafac decomposition on the generated tensor. The -o just stores whether or not to generate an output file.

These variables can start being used in the rest of the code now. I'm in the process of figuring out where to start using them in place of other variables that are used right now.

sumanthneerumalla commented 6 years ago

Need to change the decomposition argument so that its in the form of

-decom [parafac, tucker, etc..]

MaxMcGlinnPoole commented 6 years ago

I have completed this. Updates are documented in the README. the -h option should also help explain how the command line args work.