nnet models

The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.

nnet models

Function	Works
`tidypredict_fit()`, `tidypredict_sql()`, `parse_model()`	✔
`tidypredict_to_column()`	✔
`tidypredict_test()`	✔
`tidypredict_interval()`, `tidypredict_sql_interval()`	✗
`parsnip`	✔

nnet::nnet() fits feed-forward neural networks with a single hidden layer. Each hidden unit is the weighted sum of the predictors passed through the logistic squashing function, and the output units are the weighted sums of the hidden units, left unsquashed when the network was fit with linout = TRUE.

Classification models predict one probability per outcome class, so tidypredict_fit() returns a named list of expressions rather than a single expression. For those models tidypredict_to_column() and tidypredict_test() are not supported.

`tidypredict_` functions

library(nnet)

set.seed(100)
model <- nnet(mpg ~ wt + hp, data = mtcars, size = 3, linout = TRUE, trace = FALSE)

Create the R formula

tidypredict_fit(model)
#> 9.63217435895345 + case_when(-0.255509663289361 + wt * -0.309449576415116 + 
#>     hp * 0.949425509730813 < -15 ~ 0, -0.255509663289361 + wt * 
#>     -0.309449576415116 + hp * 0.949425509730813 > 15 ~ 1, .default = 1/(1 + 
#>     exp(-(-0.255509663289361 + wt * -0.309449576415116 + hp * 
#>         0.949425509730813)))) * 10.1885081177865 + case_when(-0.385719808388316 + 
#>     wt * 0.522807343827805 + hp * 20.0552277393739 < -15 ~ 0, 
#>     -0.385719808388316 + wt * 0.522807343827805 + hp * 20.0552277393739 > 
#>         15 ~ 1, .default = 1/(1 + exp(-(-0.385719808388316 + 
#>         wt * 0.522807343827805 + hp * 20.0552277393739)))) * 
#>     0.269942523260141 + case_when(0.42285989779505 + wt * -0.212906496087752 + 
#>     hp * -0.880839556387571 < -15 ~ 0, 0.42285989779505 + wt * 
#>     -0.212906496087752 + hp * -0.880839556387571 > 15 ~ 1, .default = 1/(1 + 
#>     exp(-(0.42285989779505 + wt * -0.212906496087752 + hp * -0.880839556387571)))) * 
#>     10.480592173546

Add the predictions to the original table

library(dplyr)

mtcars %>%
  tidypredict_to_column(model) %>%
  glimpse()
#> Rows: 32
#> Columns: 12
#> $ mpg  <dbl> 21.0, 21.0, 22.8, 21.4, 18.7, 18.1, 14.3, 24.4, 22.8, 19.2, 17.8,…
#> $ cyl  <dbl> 6, 6, 4, 6, 8, 6, 8, 4, 4, 6, 6, 8, 8, 8, 8, 8, 8, 4, 4, 4, 4, 8,…
#> $ disp <dbl> 160.0, 160.0, 108.0, 258.0, 360.0, 225.0, 360.0, 146.7, 140.8, 16…
#> $ hp   <dbl> 110, 110, 93, 110, 175, 105, 245, 62, 95, 123, 123, 180, 180, 180…
#> $ drat <dbl> 3.90, 3.90, 3.85, 3.08, 3.15, 2.76, 3.21, 3.69, 3.92, 3.92, 3.92,…
#> $ wt   <dbl> 2.620, 2.875, 2.320, 3.215, 3.440, 3.460, 3.570, 3.190, 3.150, 3.…
#> $ qsec <dbl> 16.46, 17.02, 18.61, 19.44, 17.02, 20.22, 15.84, 20.00, 22.90, 18…
#> $ vs   <dbl> 0, 0, 1, 1, 0, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0,…
#> $ am   <dbl> 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0, 0,…
#> $ gear <dbl> 4, 4, 4, 3, 3, 3, 3, 4, 4, 4, 4, 3, 3, 3, 3, 3, 3, 4, 4, 4, 3, 3,…
#> $ carb <dbl> 4, 4, 1, 1, 2, 1, 4, 2, 2, 4, 4, 3, 3, 3, 4, 4, 4, 1, 2, 1, 1, 2,…
#> $ fit  <dbl> 20.09063, 20.09063, 20.09063, 20.09063, 20.09063, 20.09063, 20.09…

Confirm that the results match the model’s predict() results

tidypredict_test(model, mtcars)
#> tidypredict test results
#> Difference threshold: 1e-12
#> 
#>  All results are within the difference threshold

Classification

set.seed(100)
cls_model <- nnet(Species ~ ., data = iris, size = 3, trace = FALSE)

fit <- tidypredict_fit(cls_model)
names(fit)
#> [1] "setosa"     "versicolor" "virginica"

probs <- sapply(fit, \(f) rlang::eval_tidy(f, iris))
all.equal(unname(probs), unname(predict(cls_model, iris, type = "raw")))
#> [1] TRUE

parsnip

parsnip fitted models are also supported by tidypredict:

library(parsnip)

set.seed(100)
p_model <- mlp(mode = "regression", hidden_units = 3, epochs = 100) %>%
  set_engine("nnet") %>%
  fit(mpg ~ wt + hp, data = mtcars)

tidypredict_fit(p_model)
#> 9.63217435895345 + case_when(-0.255509663289361 + wt * -0.309449576415116 + 
#>     hp * 0.949425509730813 < -15 ~ 0, -0.255509663289361 + wt * 
#>     -0.309449576415116 + hp * 0.949425509730813 > 15 ~ 1, .default = 1/(1 + 
#>     exp(-(-0.255509663289361 + wt * -0.309449576415116 + hp * 
#>         0.949425509730813)))) * 10.1885081177865 + case_when(-0.385719808388316 + 
#>     wt * 0.522807343827805 + hp * 20.0552277393739 < -15 ~ 0, 
#>     -0.385719808388316 + wt * 0.522807343827805 + hp * 20.0552277393739 > 
#>         15 ~ 1, .default = 1/(1 + exp(-(-0.385719808388316 + 
#>         wt * 0.522807343827805 + hp * 20.0552277393739)))) * 
#>     0.269942523260141 + case_when(0.42285989779505 + wt * -0.212906496087752 + 
#>     hp * -0.880839556387571 < -15 ~ 0, 0.42285989779505 + wt * 
#>     -0.212906496087752 + hp * -0.880839556387571 > 15 ~ 1, .default = 1/(1 + 
#>     exp(-(0.42285989779505 + wt * -0.212906496087752 + hp * -0.880839556387571)))) * 
#>     10.480592173546

Note that parsnip runs the class probabilities of predict.nnet() through a second softmax, so the expressions returned for a classification mlp() model match predict(model, type = "prob") rather than predict(model$fit, type = "raw").

Parse model spec

Here is an example of the model spec:

pm <- parse_model(model)
str(pm, 2)
#> List of 3
#>  $ general:List of 6
#>   ..$ model    : chr "nnet"
#>   ..$ version  : num 2
#>   ..$ type     : chr "nnet"
#>   ..$ n_units  : int 7
#>   ..$ n_outputs: num 1
#>   ..$ softmax  : logi FALSE
#>  $ inputs :List of 2
#>   ..$ :List of 2
#>   ..$ :List of 2
#>  $ units  :List of 4
#>   ..$ :List of 3
#>   ..$ :List of 3
#>   ..$ :List of 3
#>   ..$ :List of 3
#>  - attr(*, "class")= chr [1:3] "parsed_model" "pm_nnet" "list"

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.

nnet models

tidypredict_ functions

Classification

parsnip

Parse model spec

`tidypredict_` functions