search

M-Nauta / TCDF

Temporal Causal Discovery Framework (PyTorch): discovering causal relationships between time series
GNU General Public License v3.0
482 stars 106 forks source link

How to plot the predicted time series ? / How to print time delays in cause and effect ? (new to python) #11

Closed zzadiues3338 closed 3 years ago

zzadiues3338 commented 3 years ago

Howdy , I am new to python and found your code extremely useful and I am trying to implement to my dataset.

I went through the documents and it is not clear to me how to plot the predicted time series ?

additionally, how can I print the computed values of time delays in cause and effect?

Thank you very much for your time and consideration.

Sincerely, Aditya Chakravarty

M-Nauta commented 3 years ago

Hi Aditya, As shown in the demo file, the time series can be plotted by running evaluate_predictions_TCDF.py with argument --plot. The time delays between cause and effect are automatically printed when running runTCDF.py. To run TCDF on your own dataset, you simply have to set the path to the csv file as --data argument. Since you're new to python, I would recommend to do a python tutorial if you want to adapt the code to do something else. Good luck! :)