X.average.crossmodels

Average cross-trained models to get final prediction values for one dataset.

Usage

X.average.crossmodels(
  data = NULL,
  scores.col = NA,
  labels.col = NA,
  evaluate = FALSE,
  standard.set = NULL,
  eval.metric = "prc"
)

Arguments

data

Data frame with columns for protein1, protein2 and columns 'score', 'cycle' and 'split'.

scores.col

Character string: Name of the columns with prediction values.

labels.col

Character string: Name of the columns with labels.

evaluate

Logical: Should the averaged predictions be evaluated? Default is FALSE. If TRUE, standard.set must be included

standard.set

Dataframe with columns protein1, protein2 and another column with labels.

eval.metric

Character string: Evaluation metric to be used for evaluating the model performance. Options same as in X.evaluate.

Value

A list with four elements: $data is the data.frame with averaged predictions, $eval.metric is the integer specifying the resulting evaluation metric, $metric.plotdata are underlying data for plotting the evaluation metric and $plot is the plot itself.

Examples

comp.models[[rep]] <- X.average.crossmodels(data=cross.model$data, eval.metric="prc", evaluate=TRUE, standard.set=GS_specific)
#> Error: object 'cross.model' not found