Feedback for python filename classification code Jupyter notebook

[knappett]

Generally the notebook is very well laid out and ok to follow. Couple of questions below though:

1. In the fifth code block you have the following:
   one_rec_df = ff.filename_to_df("hello_or_other.nc")
   However, it's not clear why you're doing this step. Is it just to demonstrate how filename_to_df works with a dummy filename, or does it have some other purpose?
2. In the section 'What does our transformed dataset look like?' it says 'There are 1,739 columns, because we have one-hot-encoded the data' - Why are there 1739 columns after one-hot-encoding? I thought this method effectively just turned each of the possible feature values into a column, so for the feature values you specified at the start, you'd end up with about 45 columns. Perhaps I haven't understood one-hot-encoding?! This step could do with further explanation.

[agstephens]

Thanks @knappett, very useful feedback. I'm going to update and then you can review and let me know if it makes sense.

[agstephens]

@knappett, please review my changes at:
https://github.com/cedadev/ai-mini-projects/blob/main/filename-classifier/filename-classification.ipynb