How can I visualize insight in 2 predictors from a NN for a specific range of a third predictor

I have fitted a neuralnet based on 3 variables: x1, x2, x3 generating predictions y.
I would like to construct a matrix based on x1, x3 and the average prediction y for each combination while x2 is within range 0-25. (next I would like to make a similar matrix for x2 in range 26-50, 51-75 and 76-100.

I have made many attempts with dplyr and solutions based on pivottables in R although I do not succeed.

Currently I have the code below:

library(nnet)
library(caret)
library(dplyr)

x <- mydata[,2:4]
y <- mydata[,5]

parti <- createDataPartition(y, times = 1, p=0.8, list = FALSE)

x_train <- x[parti,]
x_test <- x[-parti,]
y_train <- y[parti]
y_test <- y[-parti]

fit <- nnet(y_train~., x_train, size=12, maxit=500, linout=T, decay=0.01)

x1 <- seq(0,100,10)
x2 <- seq(0,100,10)
x3 <- seq(0,100,10)

my_grid <- expand.grid(x1=x1, x2=x2, x3=x3)

predictions <- predict(fit ,my_grid, type="raw")
testResults <- data.frame(my_grid, y = predictions)

plot(testResults)

myMatrix <- testResults %>% filter(x2>0 & x2<25) %>% group_by(x1) %>% group_by(x3) %>% summarize(y=average(y))


This code generates the folowing table:

A tibble: 11 x 2
x3 y
<dbl> <dbl>
1 0 18.5
2 10 -19.2
3 20 -2.93
4 30 10.4
5 40 10.9
6 50 4.42
7 60 0.511
8 70 0.0232
9 80 -3.67
10 90 -7.26
11 100 -8.37


Although the outcome I am looking for is something like

x3 ->
x1 0 10 30 40 50 60 70 80 90 100
1 0 18.5 12 7 5
2 10 -19.2 1 3 2
3 20 -2.93 22 1 etc
4 30 10.4 3 7
5 40 10.9 4 3
6 50 4.42 5 2
7 60 0.511 3 1
8 70 0.02324 9
9 80 -3.67 5 2
10 90 -7.26 5 5
11 100 -8.37 -1 0


Based on this matrix I would like to generate a heatmap.

Thanks!

1 Answer
1

vb = testResults[testResults$x2 > 0 & testResults$x2 <= 25,]
vb = dcast(vb, x1~x3, value.var=y, fun.aggregate=mean)


Thanks for contributing an answer to Stack Overflow!

But avoid …

To learn more, see our tips on writing great answers.

Some of your past answers have not been well-received, and you're in danger of being blocked from answering.

Please pay close attention to the following guidance:

But avoid …

To learn more, see our tips on writing great answers.

Required, but never shown

Required, but never shown

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.