Look at the function stepclass() in package klaR. ------------------------------------- David L Carlson Department of Anthropology Texas A&M University College Station, TX 77840-4352
-----Original Message----- From: R-help [mailto:r-help-boun...@r-project.org] On Behalf Of David Moskowitz Sent: Sunday, February 15, 2015 11:34 AM To: n omranian via R-help Subject: [R] Picking Best Discriminant Function Variables Is there a way to have the LDA function give me the best 3 (or 4) predictor variables. When I put in all the variables, LDA uses all the variables, but I would like to know what would be the 3 (or 4) best to use out all the available variables and the coefficients for those. Here is the code I am using for Linear Discriminant Function library("MASS") results <- lda(data$V1 ~ data$V2 + data$V3 + data$V4 + data$V5 + data$V6 + data$V7 + data$V8 + data$V9 + data$V10 + data$V11 + data$V12 + data$V13 + data$V14) Output: Coefficients of linear discriminants: LD1 LD2 data$V2 -0.403399781 0.8717930699 data$V3 0.165254596 0.3053797325 data$V4 -0.369075256 2.3458497486 data$V5 0.154797889 -0.1463807654 data$V6 -0.002163496 -0.0004627565 data$V7 0.618052068 -0.0322128171 data$V8 -1.661191235 -0.4919980543 data$V9 -1.495818440 -1.6309537953 data$V10 0.134092628 -0.3070875776 data$V11 0.355055710 0.2532306865 data$V12 -0.818036073 -1.5156344987 data$V13 -1.157559376 0.0511839665 data$V14 -0.002691206 0.0028529846 So in the above example, I would like the LDA to return to me the 3 best predictors out of the 13 available. Thank you [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code. ______________________________________________ R-help@r-project.org mailing list -- To UNSUBSCRIBE and more, see https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.