pred_sens: Predicted values for Lek profile method
In NeuralNetTools: Visualization and Analysis Tools for Neural Networks

Description Usage Arguments Details Value See Also Examples

Get predicted values for Lek Profile method, used iteratively in lekprofile

1	pred_sens(mat_in, mod_in, var_sel, step_val, grps, ysel)

`mat_in`	`data.frame` of only the explanatory variables used to create model
`mod_in`	any model object with a predict method
`var_sel`	chr string of explanatory variable to select
`step_val`	number of values to sequence range of selected explanatory variable
`grps`	matrix of values for holding explanatory values constant, one column per variable and one row per group
`ysel`	chr string of response variable names for correct labelling

Gets predicted output for a model's response variable based on matrix of explanatory variables that are restricted following Lek's profile method. The selected explanatory variable is sequenced across a range of values. All other explanatory variables are held constant at the values in grps.

A list of predictions where each element is a data.frame with the predicted value of the response and the values of the explanatory variable defined by var_sel. Each element of the list corresponds to a group defined by the rows in grps at which the other explanatory variables were held constant.

lekprofile

## using nnet

library(nnet)

data(neuraldat) 
set.seed(123)

mod <- nnet(Y1 ~ X1 + X2 + X3, data = neuraldat, size = 5)

mat_in <- neuraldat[, c('X1', 'X2', 'X3')]
grps <- apply(mat_in, 2, quantile, seq(0, 1, by = 0.2))

pred_sens(mat_in, mod, 'X1', 100, grps, 'Y1')