adkinsty/post_error_followup

Follow-up analyses for the post error forced response experiments

post_error_followup

directory structure

tbd

training.ipynb

used to fit the models

inference.ipynb

used to evaluate and post-process the models

input

raw_data

data taken from dropbox, called dat.csv under each experiment subfolder

model_input

data given as input to a stan model

preprocessors

subject id encoders can be used to map between prolific ids and integer ids used in the model

output

model_fit

stan code and posterior samples for fitted models (not added yet, need to compress first)

group_par

tables and figures of group-level parameters of interest

group_pred

tables and figures of predicted behavior from group-level parameters

metrics

tables of goodness-of-fit metrics for each model (currently just WAIC)

ref

reference stan code file from dropbox used in the training notebook above.