train_valid_test_split Splits feature and target data frames into random train, validation and test subsets The proportion of the train set relative to the input data will be valid_size * (1 - test_size)

Source: R/train_valid_test_split.R

train_valid_test_split Splits feature and target data frames into random train, validation and test subsets The proportion of the train set relative to the input data will be valid_size * (1 - test_size)

train_valid_test_split(
  x,
  y,
  test_size = 0.25,
  valid_size = 0.25,
  shuffle = TRUE,
  random_state = NULL
)

Arguments

x,

y arrays
y

(data frame) Original labels from data set
test_size

(double, default = NULL) Value between 0.0 and 1.0 to represent the proportion of the dataset to comprise the size of the test subset
valid_size

float or None (default = 0.25) Value between 0.0 and 1.0 to represent the proportion of the dataset to comprise the size of the test subset
random_state

(int, default = NULL) Seed for the random number generator
x

(data frame) Original features from data set

Value

Examples

x = data.frame('X1'=c(0,1,2,3,4,5,6,7),
               'X2'=c(8,9,10,11,12,13,14,15))

y = data.frame('Y'=c(0,1,2,3,4,5,6,7))


train_valid_test_split(x,y)$x_train
#>   X1 X2
#> 8  7 15
#> 5  4 12
#> 2  1  9
#> 3  2 10
train_valid_test_split(x,y)$x_valid
#>   X1 X2
#> 7  6 14
#> 2  1  9
train_valid_test_split(x,y)$x_test
#>   X1 X2
#> 5  4 12
#> 8  7 15
train_valid_test_split(x,y)$y_train
#>   y_train
#> 1       6
#> 2       1
#> 3       0
#> 4       7
train_valid_test_split(x,y)$y_valid
#>   y_valid
#> 1       6
#> 2       0
train_valid_test_split(x,y)$y_test
#>   y_test
#> 1      6
#> 2      4