Function to calculate and plot the accuracy, hinge loss and binary cross-entropy loss of a machine learning prediction using the ppi.prediction function.

learning.curve(
  ppi_prediction_result,
  train_sizes = base::seq(0.1, 1, by = 0.1),
  models = "all",
  verbose = TRUE,
  prob = 0.5,
  sum_data = "Median"
)

Arguments

ppi_prediction_result:

result from the function ppi.prediction

train_sizes:

sequence of fraction of training sizes to be used for calculation from >0 to 1

models:

Integer of models used to calculate the loss functions. If "all" then all models as specified by the ensembleSize in ppi.prediction will be used.

verbose:

boolean, prints detailed informations

prob:

probability cutoff to calculate accuracy

sum_data:

summary statistic to plot: 'Mean', 'Median', or 'None'

Value

a list with elements

Examples

data("example_ppi_prediction")
plot <- learning.curve(example_ppi_prediction)
#> learning curve for model 1 completed.
#> learning curve for model 2 completed.
#> learning curve for model 3 completed.
#> learning curve for model 4 completed.
#> learning curve for model 5 completed.
#> learning curve for model 6 completed.
#> learning curve for model 7 completed.
#> learning curve for model 8 completed.
#> learning curve for model 9 completed.
#> learning curve for model 10 completed.
#> learning curve for model 11 completed.
#> learning curve for model 12 completed.
#> learning curve for model 13 completed.
#> learning curve for model 14 completed.
#> learning curve for model 15 completed.
#> learning curve for model 16 completed.
#> learning curve for model 17 completed.
#> learning curve for model 18 completed.
#> learning curve for model 19 completed.
#> learning curve for model 20 completed.
#> learning curve for model 21 completed.
#> learning curve for model 22 completed.
#> learning curve for model 23 completed.
#> learning curve for model 24 completed.
#> learning curve for model 25 completed.
#> learning curve for model 26 completed.
#> learning curve for model 27 completed.
#> learning curve for model 28 completed.
#> learning curve for model 29 completed.
#> learning curve for model 30 completed.
#> learning curve for model 31 completed.
#> learning curve for model 32 completed.
#> learning curve for model 33 completed.
#> learning curve for model 34 completed.
#> learning curve for model 35 completed.
#> learning curve for model 36 completed.
#> learning curve for model 37 completed.
#> learning curve for model 38 completed.
#> learning curve for model 39 completed.
#> learning curve for model 40 completed.
#> learning curve for model 41 completed.
#> learning curve for model 42 completed.
#> learning curve for model 43 completed.
#> learning curve for model 44 completed.
#> learning curve for model 45 completed.
#> learning curve for model 46 completed.
#> learning curve for model 47 completed.
#> learning curve for model 48 completed.
#> learning curve for model 49 completed.
#> learning curve for model 50 completed.