Run process.R to convert the raw data from the SPSS file into an R dataframe (note: original SPSS file is not in this repo). This will also generate other useful files in the workspace folder.
To get Alspac sleep variables, run prepare.R in the sleep folder. This will create a dataframe for sleep data. The file learn.py can be used to learn a State Space Model using PyStan based on the timeseries sleep data. The program_miss.stan file describes the model (which can handle missing data). The file vb_learn.py is identical to learn.py but uses variational inference instead of MCMC.
To setup the social behaviour data, successively run prepare_1.py and prepare_2.R (sorry). The vae_learn.py script can be used to learn a VAE model on a single time point of the social behaviour timeseries. The train.py script can be used to train a Sequential VAE model on the entire timeseries. It can also be used to train the model on simulated data. Models are built in tensorflow and training is done using the tf.Estimator class.
The var() method in search.py can be used to search for variable names within the dataframe.