-
Notifications
You must be signed in to change notification settings - Fork 5
/
RankRSA.m
29 lines (29 loc) · 1.03 KB
/
RankRSA.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
% Spearman rho-based representational similarity analysis. See also the RSA,
% GLM superclasses.
%
% gl = RankRSA(modelrdms,datardms)
classdef RankRSA < RSA
methods
function gl = RankRSA(modelrdms,datardms)
if nargin==0
modelrdms = [];
datardms = [];
end
% use super-class to initialise
gl = gl@RSA(modelrdms,datardms);
% then rank trans and Z score so that linear fits become
% equivalent to Spearman rho
gl.X = zscore(ranktrans(gl.X),0,1);
gl.data = zscore(ranktrans(gl.data),0,1);
% finally reduce precision since we are working on ranked data
% anyway (you'd need a very large number of unique ranks to run
% into any precision trouble with single floats)
if isa(gl.X,'double')
gl.X = single(gl.X);
end
if isa(gl.data,'double')
gl.data = single(gl.data);
end
end
end
end