n-gram range in textacy.extract.ngrams

[mzeidhassan]

I have a question about "textacy.extract.ngrams". I am trying to extract ngrams in the range of (2, 3, 4, 5) grams. I tried to pass the following, but it doesn't seem to work.

ngrams = textacy.extract.ngrams(doc, n=(2, 6), filter_stops=True, filter_nums=True, min_freq=2)

Is there a way to pass a range of n-grams? Or is there a better module that can do what I am looking for?

Please note that I am also using "exclude_pos", that's why I opted for 'extract.ngrams'.

Environment

• operating system: Windows 10
• python version: 3.6
• spacy version: 2.1.8
• installed spacy models: 'en_core_web_lg'
• textacy version: 0.9.1

Thanks

[bdewilde]

Hi @mzeidhassan , extract.ngrams() only accepts a single integer for n, so you'd have to loop over the values and concatenate the results. Something like this would do it:

ngrams = [
    ngram
    for n in range(2, 6)
    for ngram in extract.ngrams(doc, n, filter_stops=True, filter_nums=True, min_freq=2)
]

[mzeidhassan]

Thanks @bdewilde for your support.

I just want to say thank you for creating such a great library. Textacy is really amazing. Wish you all the best!

Thanks, Mohamed