Missing data
Handling missing data is a critical step in Data cleaning. Understanding missing data is critical because the missingness is not always completely random. Ideally, we need to understand the reasons for missingness (or model the missingness).
- Missing-data imputation by Andrew Gelman from http://www.stat.columbia.edu/~gelman/arm/
- Missing data by Gary King.
Methods
- Multiple imputation
- You2020handling suggests a method using Graph embedding
- https://amices.org/mice/
- https://cran.r-project.org/web/packages/brms/vignettes/brms_missings.html