Utilities
These utility functions are meant to provide easy comprehended overall evaluation of Models and Algorithms. For now, the support is limited to Classification Models, but extended further in future.
Metrics.report_stats
— Functionreport_stats(y_pred, y_true; classwise_stats=true, avg_type="macro", sample_weights=nothing)
A utility function that prints the statistics summary of the model based on provided y_pred
and y_true
.
Arguments:
y_pred
: predicted valuesy_true
: ground truth values on the basis of which predicted values are to be assessed.classwise_stats=true
: if settrue
, prints classwise stats along with global stats.avg_type="macro"
: Type of average to be used while calculating precision of multiclass models. Can take values asmacro
,micro
andweighted
. Default set tomacro
.sample_weights
: Class weights to be provided whenavg_type
is set toweighted
. Useful in case of imbalanced classes.