SPOTMisc (version 1.19.52)

getMnistData: getMnistData

Description

Based on the setting kerasConf$encoding either one-hot encoded data or tensor-shaped data are returned.The labels are converted to binary class matrices using the function to_categorical.

Usage

getMnistData(kerasConf)

Value

list with training and test data, i.e., list(x_train, x_test, y_train, y_test).

Arguments

kerasConf

List of additional parameters passed to keras as described in getKerasConf. Default: NULL.

See Also

getKerasConf

funKerasMnist

Examples

# \donttest{
### These examples require an activated Python environment as described in
### Bartz-Beielstein, T., Rehbach, F., Sen, A., and Zaefferer, M.:
### Surrogate Model Based Hyperparameter Tuning for Deep Learning with SPOT,
### June 2021. http://arxiv.org/abs/2105.14625.
PYTHON_RETICULATE <- FALSE
if(PYTHON_RETICULATE){

library("SPOTMisc")
kerasConf <- getKerasConf()
kerasConf$encoding <- "oneHot" # default
mnist <- getMnistData(kerasConf)
# lots of zeros, but there are also some nonzero (greyscale) values, e.g.:
mnist$x_train[1,150:160]
str(mnist$x_train[1,])
# y-labels are one-hot encoded. The first entry represents "5"
mnist$y_train[1,]
##
kerasConf$encoding <- "tensor"
mnist <- getMnistData(kerasConf)
## 28x28:
str(mnist$x_train[1,,,])
mnist$y_train[1,]
}
# }