Plots Accumulated Local Effects for ERF for multivariate factors — plot_ALE

Plots the Accumulated Local Effects (ALE) from an ERF object

plot_ALE_multi(
  ALE,
  xquantiles = c(0.025, 0.975),
  yquantiles = c(0.1, 0.5, 0.9),
  name,
  cex.axis = 1,
  cex.lab = 1,
  rug = TRUE,
  rug.col = "gray50",
  xaxs = TRUE,
  gap.axis = NA,
  rug.tick = 0.02,
  rug.lwd = 0.5,
  rug.alpha = 0.2,
  rug.max = 1000,
  poly.colp = c("#fcde9c", "#e34f6f", "#7c1d6f"),
  poly.alpha = 0.5,
  level.names
)

Arguments

ALE: an ALE object for a given variable; indexed as a list for full functionality
xquantiles: lower and upper quantile bounds to limit x values to
yquantiles: lower, middle, and upper quantiles to plot the confidence interval around the ALE predictions
name: the x-axis label of the variable; self-supplied if indexed as a list
cex.axis: the cex of the axis
cex.lab: the cex of the labels
rug: logical for turning the rug plot on/off (defaults to table values for class=='factor')
rug.col: color of rug ticks
rug.tick: length of rug ticks
rug.lwd: width of rug ticks
rug.alpha: transparency of rug ticks
rug.max: maximum number of ticks to plot (useful for big data)
poly.colp: is a vector of colors equal to n levels in response metric
poly.alpha: is a transparency value between 0 and 1 for the confidence interval polygon from the ERF predictions
level.names: names of the multivariate levels

Value

A plot of the ALEs

Examples

#run an ERF with 10 RFs and 
logit <- function(x){log(x/(1-x))}
inv_logit <- function(x){exp(x)/(exp(x)+1)}

x_mat <- as.data.frame(replicate(4, rnorm(1e4)))
x_mat_bin <- t(rmultinom(1e4,1,prob=c(0.33,0.33,0.33)))
x_mat$V5 <- factor(apply(x_mat_bin,1,function(x)which(x==1)))

y_gin <- function(){
  eff <- rnorm(ncol(x_mat)-1)
eff_bin <- rnorm(ncol(x_mat_bin))
y_logit <- t(eff %*% t(x_mat[,-5])) + t(eff_bin %*% t(x_mat_bin))
y <- inv_logit(y_logit)
return(y)
}

y_poss <- data.frame(y1 = y_gin(),
                    y2 = y_gin(),
                    y3 = y_gin())

y_bin <- apply(y_poss,1,function(x) rmultinom(1,1,x))
y <- apply(y_bin,2,function(x)which(x==1))
df <- data.frame(y = factor(y),x_mat)

ens_rf_ex <- ens_random_forests(df=df, var="obs", covariates=colnames(df[,-1]), save=FALSE, cores=1)
#> rounding n.forests to the nearest one
#> Error in `[.data.frame`(df, , var): undefined columns selected


ALE_df <- ALE_fn(ens_rf_ex, save=FALSE, multi=TRUE, type='prob')
#> Error in ALE_fn(ens_rf_ex, save = FALSE, multi = TRUE, type = "prob"): unused argument (save = FALSE)

plot_ALE_multi(ALE_df[1])
#> Error in eval(expr, envir, enclos): object 'ALE_df' not found