Median Polish

Fits an additive model using Tukey's median polish procedure.

MEDIAN_POLISH(X) fits an additive model to a 2D input matrix, X, assuming a decomposition of (constant + rows + columns).

Median polishing is useful for removing spatial trends in the data by alternately removing medians from the rows and columns of the data. The algorithm proceeds as follows. Considering the rows first, for each row the row median is subtracted from every element in that row. Then, for each column, the median of the revised numbers is then subtracted from every element in that column. This process is repeated until all convergence, or a maximum number of iterations.