MAR FunctionsΒΆ

Missing At Random (MAR) refers to a missing data mechanism where the probability of missingness depends only on the observed data.

Formally, let \(X\) be the data matrix and \(M\) the missingness indicator. Then under MAR:

\[P(M \mid X) = P(M \mid X_{\text{obs}})\]

This implies that the probability of missingness may depend on observed variables, but not on the values that are missing themselves.

In MissMecha, we provide multiple MAR variants for both numerical and categorical data. These functions are especially useful for simulating realistic missingness patterns in applied data science tasks.

Note

MAR mechanisms support all data types, including numerical, categorical, and time series inputs (when treated as tabular data).

Numerical MAR Types

Numerical Types Suitable for continuous or ordinal variables. Includes logistic regression masking, correlation-based ranking, and extreme-value-based strategies.

Categorical MAR Types

Categorical Types Designed for discrete or binary variables. Supports group-based missingness and masking driven by feature associations.