multitrait.plot: Accuracy vs penalization from Multitrait SSI

In SFSI: Sparse Family and Selection Index

Accuracy vs penalization from Multitrait SSI

Description

Visualizing results from an object of the class 'SSI'

Usage

multitrait.plot(object, x.stat = c("nsup","lambda"),
                y.stat = c("accuracy","MSE"),
                line.color = "orange",
                nbreaks.x = 7, ...)

Arguments

object	An object of the class 'SSI' for a multitrait case
x.stat	(character) Either 'nsup' (number of non-zero regression coefficients entering in the prediction of a given testing individual) or 'lambda' (penalization parameter in log scale) to plot in the x-axis
y.stat	(character) Either 'accuracy' (correlation between observed and predicted values) or 'MSE' (mean squared error) to plot in the y-axis
line.color	(character) Color of the lines
nbreaks.x	(integer) Number of breaks in the x-axis
...	Other arguments for method plot: 'xlab', 'ylab', 'main', 'lwd', 'xlim', 'ylim'

Value

Creates a plot of either accuracy or MSE versus either the support set size (average number of predictors with non-zero regression coefficient) or versus lambda. This is done separately for each trait

Examples

  require(SFSI)
  data(wheatHTP)
  
  index = which(Y$trial %in% 1:9)   # Use only a subset of data
  Y = Y[index,]
  M = scale(M[index,])/sqrt(ncol(M))  # Subset and scale markers
  G = tcrossprod(M)                   # Genomic relationship matrix
  y = scale(Y[,4:5])    # Response variable
  
  # Sets (testing=0, training=1)
  trn_tst = matrix(NA,ncol=ncol(y),nrow=nrow(y))
  trn_tst[,1] = ifelse(Y$trial %in% 2, 0, 1)
  trn_tst[,2] = ifelse(Y$trial %in% 2, 0, 1)
  
  fm = SSI(y, K=G, trn_tst=trn_tst, mc.cores=1)
  multitrait.plot(fm)
  

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