Code to train a word2vec vector-space model on the MythFic corpus (not included), plot emotion associations per character (see /radarplots/), emotion associations per character comparing different genres (see /genreradarplots/), and distribution of emotion associations (over a larger set of characters) differentiated by gender and genre (see /violinplots/)
pip install -r requirements.txt
Separately also make sure you have sklearn.
MythFic_txt directory containing [id].txt files should be at the same level as this repository (i.e. one level up from this README), as well as the file fanfics_Greek_myth_metadata.csv available here. Then from within myth2vec just run python myth2vec.py. Training takes a while so the model is cached using pickle; therefore to retrain (after changing corpus or model parameters) remove all pickle files.