>Where would one modify this code to filter different sentiments?

Right here: https://github.com/garritfra/positive_hackernews/blob/main/m...

This project uses the sentiment analyzer provided by the NLTK toolkit [0]. The `SentimentIntensityAnalyzer` class assigns polarity scores to its text input, where a score >=0 denotes a "positive valence".

[0]:https://www.nltk.org/api/nltk.sentiment.SentimentIntensityAn...