Building neural network fails on test cases outside BIOME-z example RDF
jtanderson opened this issue · 0 comments
jtanderson commented
Describe the bug
The parsing/use of other RDF libraries throws an error. This seems to be because there are assumptions imposed on the structure of the parsed library.
Relevant log information:
File "/home/jtanderson/Documents/BIOME-z-Project/GUI/datasets.py", line 24, in _csv_iterator
yield int(row[0]) - 1, ngrams_iterator(tokens, ngrams)
ValueError: invalid literal for int() with base 10: 'Covid-19’s disruption of India’s transformed food supply chains'
To Reproduce
Steps to reproduce the behavior:
- Import the "B-RaPIDD2" RDF file currently added to the
GUI
folder for testing. - Attempt to build the neural network