Evaluate model performance for binary tasks
Usage
performance.metrics(actual, pred, return_contingency = FALSE)
Arguments
- actual
Logical or numeric binary vector giving the actual cell labels.
- pred
Logical or numeric binary vector giving the predicted cell labels.
- return_contingency
Logical indicating if contingency table must be returned.
Value
Prediction performance metrics (Precision, Recall, MCC) between actual and
predicted cell type labels.
Examples
results <- performance.metrics(actual= sample(c(1,0),20,replace=TRUE),
pred = sample(c(1,0),20,replace=TRUE,prob = c(0.65,0.35) ) )