Parse_data_from_file only finding 18 sentences

Course Q&A TensorFlow Developer Professional Natural Language Processing in TensorFlow
1

Good evening,

Currently experiencing a few issues and appreciate any advice!

  1. My parse_data_from_file function uses a for loop, which, on testing, only finds 18 sentences in the original dataset. It then fails with a,

‘list’ object has no attribute ‘split’

error on the ‘print(f"First sentence has {len(sentences[0]…’ line. Slightly confused here!

  1. Slightly separate issue (and unrelated to the above, as I have currently removed this line). When I include my remove_stopwords function, I receive a

‘list’ object has no attribute ‘lower’

error. I included the ‘sentence = sentence.lower()’ argument in my ‘remove_stopwords’ function, as without it, the function would not remove the word ‘I’.

Notebook available for those brave enough to have a look!

2

Fixed it! I was using ‘filename’ rather than ‘csvfile’ in the csv.reader portion.

I’d also mixed up the remove_stopwords function and had some elements accidentally sat outside the function when I was moving lines of code around. D’oh!

3

What would be the input in the delimiter function?

4

There is no delimiter function. When using csv.reader, the delimiter should be set to the character on which the fields are split.