[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Unable to install Tisean

From: Juan Pablo Carbajal
Subject: Re: Unable to install Tisean
Date: Wed, 8 Aug 2018 22:48:38 +0200


If you only want PCA you do not need tisean at all.
If X is your dataset (rows: samples, cols= variables)

X_ = X - mean (X); # center variables
[U S V] = svd (X_, 1);
PCA_basis = V; # columns are your PCA vectors
P = S * U.';  # These are the scores, such that X_ = V * P
lambda = diag (S).^2 / ( size(X,1) - 1); # Eigenvalues of the sample
covaraince matrix
cumvar = cumsum (lambda) / sum (lambda); # explained variance as
function of number of components


reply via email to

[Prev in Thread] Current Thread [Next in Thread]