ResidualMatrix

ResidualMatrix provides delayed computation of a matrix of residuals after fitting a linear model to each column of an input matrix.

Key Features:

• Delayed Residual Computation: Computes residuals after fitting a linear model to each column of an input matrix without immediate realization of the full matrix.
• Partial Residual Computation: Supports computation of residuals where selected factors are preserved in the output matrix.
• Efficient Matrix Operations: Implements efficient methods for operating on the delayed matrix of residuals, including matrix multiplication and calculation of row/column sums or means.

Scientific Applications:

• Linear Model Residual Analysis: Enables extraction of residual matrices for downstream analyses while controlling for confounding variables in large-scale matrix data.
• Factor Preservation in Residuals: Allows researchers to remove unwanted variation while retaining biologically relevant factors of interest in the residual matrix.

Methodology:

Fits a linear model to each column of an input matrix and computes the residual matrix with delayed evaluation; selected factors can be preserved in the output.