This script loads all Bite reviews and uses [textblob](https://textblob.readthedocs.io/en/dev/) to see which Bites have the most negative vs. raving reviews.
It's a great example of using a library that abstracts all (NLP) complexity away and got me results fast!
Run make setup to make a virtual environment and install the dependencies.
Output of running python script.py:
$ python script.py
bite id | # comments | avg sentiment score
442 | 1 | -0.5833333333333333
142 | 3 | -0.38675925925925925
286 | 3 | -0.2138227513227513
145 | 4 | -0.16699305555555555
276 | 3 | -0.16435185185185186
260 | 6 | -0.1388888888888889
[output trucanted]
412 | 2 | 0.7375
140 | 2 | 0.75
229 | 3 | 0.7533333333333333
To read reviews for one Bite:
- Positive:
$ python script.py 229
0.6 | Nice one.
0.75 | Nice Bite! Learned (once again) to always, always, always proof-read my code.
0.91 | I've always struggled with loops, so this was very good practice.
- Negative:
$ python script.py 276
-0.25 | It was only difficult because I forgot why we were defining.
-0.15 | I'm not sure if I'm missing something on this one. the multiply_numbers function didn't require a and b to be passed in to it when it was defined. It can still multiply a and b.
-0.09 | If you are already going to have a function defined in the starting code it would be less confusing if you already had the multiply_numbers(a, b): filled in. As there is no way to get output from the built in editor, it is hard to troubleshoot that the variables in the function are missing.