Connecting the Dots Variables of Literary History and Emotions in German-language Poetry - Code and Data

Usage
Experiments can be replicated by running run.py. Please take a look at run.py, since there are some choices to be made within the code. Since pymc4 and gpu usage is under heavy development at the time, we advise to stick to the excact package versions from requirements.txt.
Content

data

input.tsv - tab-separated table containing corpus metadata (including emotions) 
factors

model.py - declaration the HGLM in pymc3
preprocessing.py - code to transform the data in input.tsv to fit the HGLM
plots
requirements.txt - required python version and packages
run.py - wrapper to start the experiments (priors and targeted emotions need to be decalered in the code)