Methods_LASSO: LASSO methods

In SFSI: Sparse Family and Selection Index

LASSO methods

Description

Retrieving regression coefficients and predicted values from the 'solveEN' and 'LARS' functions' outputs

Usage

## S3 method for class 'LASSO'
coef(object, ...)

## S3 method for class 'LASSO'
fitted(object, ...)

Arguments

object

An object of the class 'LASSO' returned either by the function 'LARS' or 'solveEN'

...

Other arguments: X (numeric matrix) scores for as many predictors there are in ncol(object$beta) (in columns) for a desired number n of observations (in rows) iy (integer vector) Optional index of columns of the matrix 'Gamma' to be returned in coef function ilambda (integer) Optional to return regression coefficients associated to a specific penalty position nsup (numeric) Optional to return regression coefficients associated to a given penalty that yield approximately 'nsup' non-zero coefficients

Value

Method coef returns a matrix that contains the regression coefficients (in rows) associated to each value of lambda (in columns). When the regression was applied to an object Gamma with more than one column, method coef returns a list

Method fitted returns a matrix with fitted values Xβ (in rows) for each value of lambda (in columns).

Examples

  require(SFSI)
  data(wheatHTP)
  
  y = as.vector(Y[,"E1"])  # Response variable
  X = scale(X_E1)          # Predictors
  
  # Training and testing sets
  tst = which(Y$trial %in% 1:10)
  trn = seq_along(y)[-tst]

  # Calculate covariances in training set
  XtX = var(X[trn,])
  Xty = cov(X[trn,],y[trn])
  
  # Run the penalized regression
  fm = solveEN(XtX,Xty,alpha=0.5)   
  
  # Regression coefficients
  dim(coef(fm))
  dim(coef(fm, ilambda=50)) # Coefficients associated to the 50th lambda
  dim(coef(fm, nsup=25))    # Coefficients where around nsup=25 are non-zero
  
  # Predicted values
  yHat1 = fitted(fm, X=X[trn,])  # training data
  yHat2 = fitted(fm, X=X[tst,])  # testing data
  
  # Penalization vs correlation
  plot(-log(fm$lambda[-1]),cor(y[trn],yHat1[,-1]),main="training",type="l")
  plot(-log(fm$lambda[-1]),cor(y[tst],yHat2[,-1]),main="testing",type="l")

SFSI documentation built on Nov. 18, 2023, 9:06 a.m.