Description Usage Arguments Value Examples
View source: R/neural_net_visualization.R
plot_partial_dependencies
creates the partial dependence plot for
the specified predictors.
1 2 3 |
neural_net |
The NeuralNetwork instance, see:
|
predictors |
Vector of predictors of the neural network for which to plot the partial dependencies. |
probs |
Vector of lower and upper bound probabilities for the confidence interval. If both are 0, intervals will not be plotted. |
type |
Either 'ggplot' if the plot should be created using 'ggplot' or 'ggplotly' if plotly should be used. |
nrepetitions |
Number of samples used within bootstrap for confidence intervals. |
parallel |
Boolean specifying if for the bootstrap confidence interval the plotting data creation should be parallelized. |
use_stored_data |
Boolean specifying if the stored data within a model should be used. Raises an error if no stored data is available. |
Created figure
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 | ## Not run:
# Example: Numeric
library(MASS)
neural_network <- NeuralNetwork(f = medv ~ ., data = Boston,
layers = c(5, 3), scale = TRUE,
linear.output = TRUE)
plot_partial_dependencies(neural_network, predictors = "crim",
probs = c(0.2, 0.8), type = "ggplotly")
plot_partial_dependencies(neural_network, predictors = c("crim", "age"))
plot_partial_dependencies(neural_network, probs = c(0.1, 0.9))
# Example: Categoric or Binary
library(datasets)
model <- NeuralNetwork(
Species ~ Sepal.Length + Sepal.Width + Petal.Length + Petal.Width,
data = iris, layers = c(10, 10), rep = 5, err.fct = "ce",
linear.output = FALSE, lifesign = "minimal", stepmax = 1000000,
threshold = 0.001, scale = FALSE)
plot_partial_dependencies(model, predictors = "Petal.Length")
plot_partial_dependencies(model,
predictors = c("Sepal.Length", "Petal.Length"))
plot_partial_dependencies(model, type = "ggplotly")
## End(Not run)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.