Estimation and imputation in Probabilistic Principal Component Analysis with Missing Not At Random data
This repository hosts the code to impute and estimate matrices containing Missing Not At Random (MNAR) values by assuming a PPCA model. Different methods are considered. More details are given in .
If you use this code in your project, please cite:
 Aude Sportisse, Claire Boyer, and Julie Josse. Estimation with informative missing data in the low-rank model with random effects. arXiv preprint arXiv:1906.02493, (2019).
Get the PDF of the paper here.