PYNLPL-COMPUTEPMI(1)

PYNLPL-COMPUTEPMI 💖

okay bestie, so this is the manual page for pynlpl-computepmi — it's a User Commands situation, version 0.7.7, and it lives in the pynlpl family. pull up a chair, we're computing co-occurrences and we're gonna have a time. 🦋

✨ NAME ✨

computepmi — the manual page for pynlpl-computepmi 0.7.7, no cap, that's literally what it says on the tin. 🎀

🌸 DESCRIPTION 🌸

so like. the usage is this, exactly this, character for character, don't come for me:

pynlpl-computepmi [-h] -f INPUTTEXT [-s] [-t THRESHOLD] [-a] [-A]

       [--pmi] [--npmi] [--jaccard] [--dice]

and what does it do? simple cooccurence computation. that's the whole pitch. that's the vibe. 🫶

🎀 optional arguments 🎀

these are the little knobs you get to twist. every single one, in order, nobody's getting left behind:

• -h, --help — shows this help message and then exits. it's giving "ask and you shall receive". 💅 (default: it just does the thing)
• -f INPUTTEXT, --inputtext INPUTTEXT — the input file, bestie. and be specific about what it wants: plaintext, tokenised, utf-8, and one sentence per line. that's the format, periodt. (default: empty, i.e. (default: )) 🙈
• -s, --sorted — output sorted by co-occurrence score. if you want your results looking all neat and ranked and main-character, this is the one. (default: False) ⭐
• -t THRESHOLD, --threshold THRESHOLD — the joined occurrence threshold. words that occur less than this? not considered. not even a little bit. it's giving gatekeeping but make it statistics. (default: 1) 🔥
• -a, --adjacency — computes the adjacency fraction, which is how many co-occurrence are immediate bigrams. yes bestie, we're counting the neighbours. (default: False) 🌈
• -A, --discountadjacency — do not take immediately adjacent fragments (bigrams) into account when computing mutual information. ⚠️ heads up: this one requires -a. (default: False) 💀
• --pmi — computes pointwise mutual information. the whole reason we're here tbh. (default: False) 👑
• --npmi — computes normalised pointwise mutual information. pmi but it went to finishing school. (default: False) 🧚
• --jaccard — computes the jaccard similarity coefficient. (default: False) 🍓
• --dice — computes the dice coefficient. (default: False) 🎲

ngl bestie, that's the whole page. every flag, every default, every threshold — accounted for. go compute something iconic. 💜

and that's a wrap! 🫶 this was pynlpl-computepmi 0.7.7, dated February 2016, and honestly? she's still slaying. bye babes ✨💖