Sensitivity analysis plot over time of the data

Plot of sensitivity of the neural network output respect to the inputs over the time variable from the data provided

Usage

CombineSens(object, comb_type = "mean")

Arguments

object: SensMLP object generated by SensAnalysisMLP with several outputs (classification MLP)
comb_type: Function to combine the matrixes of the raw_sens component of object. It can be "mean", "median" or "sqmean". It can also be a function to combine the rows of the matrixes

Value

SensMLP object with the sensitivities combined

Examples

# \donttest{
fdata <- iris
## Parameters of the NNET ------------------------------------------------------
hidden_neurons <- 5
iters <- 250
decay <- 0.1

#' ## TRAIN nnet NNET --------------------------------------------------------
# Create a formula to train NNET
form <- paste(names(fdata)[1:ncol(fdata)-1], collapse = " + ")
form <- formula(paste(names(fdata)[5], form, sep = " ~ "))

set.seed(150)
mod <- nnet::nnet(form,
                  data = fdata,
                  linear.output = TRUE,
                  size = hidden_neurons,
                  decay = decay,
                  maxit = iters)
#> # weights:  43
#> initial  value 202.395197 
#> iter  10 value 89.434352
#> iter  20 value 69.265889
#> iter  30 value 30.753230
#> iter  40 value 23.999431
#> iter  50 value 23.111929
#> iter  60 value 23.041209
#> iter  70 value 23.013926
#> iter  80 value 23.007422
#> final  value 23.007413 
#> converged
# mod should be a neural network classification model
sens <- SensAnalysisMLP(mod, trData = fdata, output_name = 'Species')
#> Loading required namespace: ggforce
#> Warning: All aesthetics have length 1, but the data has 4 rows.
#> ℹ Did you mean to use `annotate()`?
#> Warning: minimum occurred at one end of the range
#> Warning: minimum occurred at one end of the range
#> Warning: minimum occurred at one end of the range
#> Warning: minimum occurred at one end of the range

combinesens <- CombineSens(sens, "sqmean")
# }