[XLOADINGS,YLOADINGS] = plsregress(X,Y,NCOMP)
// Ncomp:主成分個數
// XLOADING : X 的線性組合系數矩陣
//YLOADING : y
// XSCORES is an N-by-NCOMP orthonormal matrix with rows corresponding to observations, columns to components.
// YSCORES is an N-by-NCOMP matrix with rows corresponding to observations,columns to components. YSCORES is neither orthogonal nor normalized.
//BETA is a (P+1)-by-M matrix, containing intercept terms(截距項) in the first row, i.e., Y = [ONES(N,1) X]*BETA + Yresiduals, and Y0 = X0*BETA(2:END,:) + Yresiduals.
// PCTVAR containing the percentage of variance explained by the model.
// MSE containing estimated mean squared errors for PLS models with 0:NCOMP components.
[XL2,YL2,XS2,YS2,BETA2,PCTVAR2,MSE2,stats2] =plsregress(a,b,ncomp)
beta3(1,:)=mu(n+1:end)-mu(1:n)./sig(1:n)*BETA2([2:end],:).*sig(n+1:end) %原始數據回歸方程的常數項
beta3([2:n+1],:)=(1./sig(1:n))'*sig(n+1:end).*BETA2([2:end],:) %計算原始變量x1,...,xn的系數，每一列是一個回歸方程
sig = std(data)
-----------------------------------------------------------------------------------------------------
Χ1' = -4.1306 * u1+0.0558 *u2t
XL2 =
-4.1306 0.0558
-4.1933 1.0239
2.2264 3.4441
YL2 =
2.1191 -0.9714
2.5809 -0.8398
0.8869 -0.1877
XSCORES is an N-by-NCOMP orthonormal matrix with rows corresponding to observations, columns to components.