Wong Chun Kit
2006-Apr-09 05:51 UTC
[R] Difference between Traditional Regression and Partial Least Square
>>Dear R-Helpers, I've a data set and run the traditional regression and partial least square as below:>lm(y~x10+x11+x12+x13+x14+x15+x16, data=X)Coefficients: Estimate Std. Error t value Pr(>|t|) (Intercept) -0.4762538 0.0252618 -18.853 < 2e-16 *** x10 0.2825081 0.0962377 2.936 0.00348 ** x11 0.0487763 0.1222990 0.399 0.69019 x12 0.0189079 0.1200368 0.158 0.87490 x13 0.0957643 0.1236650 0.774 0.43907 x14 -0.2028041 0.1243989 -1.630 0.10367 x15 -0.0005613 0.1255884 -0.004 0.99644 x16 0.0815347 0.0837342 0.974 0.33066>plsr(formula = y ~ x10 + x11 + x12 + x13 + x14 + x15 + x16, 7,data = X)y x10 0.2825080818 x11 0.0487762894 x12 0.0189078718 x13 0.0957643290 x14 -0.2028040503 x15 -0.0005613228 x16 0.0815347421 I checked that the estimated coefficient is the same. What is the difference between lm and plsr? Thanks in advance.