Open mfeblowitz opened 6 years ago
^^ Thanks! The new corefgraph may not be fully compatible with our code due to some format changes in the output. In fact we do not really need corefgraph. The current model does not have to use features generated from there.
In that case, go to code/notes/utilities/pre_processing/news_reader.py comment out L70, L73 uncomment L72
To test: start python in your TEA path, run from code.notes.utilities.pre_processing.news_reader import NewsReader print NewsReader().pre_process("this is a test")
According to https://bitbucket.org/Josu/corefgraph, should be: