peppr.Metric#

class peppr.Metric[source]#

The base class for all evaluation metrics.

The central evaluate() method takes a for a system reference and pose structures as input and returns a sclar score.

Attributes:
namestr

The name of the metric. Used for displaying the results via the Evaluator. ABSTRACT: Must be overridden by subclasses.

thresholdsdict (str -> float)

The named thresholds for the metric. Each threshold contains the lower bound

__init__() None#

Methods

__init__()

evaluate(reference, pose)

Apply this metric on the given predicted pose with respect to the given reference.

smaller_is_better()

Whether as smaller value of this metric is considered a better prediction.

Attributes

name

thresholds