reverse-shooting

Matlab scripts for reverse shooting
Log | Files | Refs | README

matrixR2.m (565B)

      1 function R2=matrixR2(A,B);
      2 % function []=matrixR2(A,B);
      3 %
      4 %  Given two matrices, A and B, of the same dimensions, compute an R2-like
      5 %  measure of how similar they are:
      6 %
      7 %      R2=1-RSS/TSS, where RSS = sum of squared deviations (element by element)
      8 %                      and TSS = total sum of squares of each element
      9 %
     10 %    Like SUM((a_ij-b_ij)^2) / SUM(a_ij ^2+b_ij^2)
     11 %
     12 %  Note, we first take out the mean value for each matrix.
     13 
     14 a=demean(vector(A)); b=demean(vector(B));  % Stack each matrix into a vector
     15 RSS=sum((a-b).^2);
     16 TSS=sum(a.^2+b.^2);
     17 R2=1-RSS/TSS;