TSGap documentation
TSGap is a Python library for composable time-series missingness simulation. It separates missingness mechanisms (why values are missing) from temporal patterns (how the missing values are arranged), making it possible to construct reproducible benchmarking scenarios for imputation methods.