base.metrics.error_metric module

class base.metrics.error_metric.ErrorMetric(default_value, max_value, step)

Bases: abc.ABC

Abstract Base Class for an error metric. Inheriting classes must implement methods:

__init__ getValueStr getNextValue getPrevValue greaterMax

Parameters:

• default_value – The default error value of the new ErrorMetric.
• max_value – The maximum error value of the new ErrorMetric.
• step – The step size, i.e. the rate at which the error value is increased.

Note

All value parameters can have arbitrary types to allow for arbitrary error definitions.

static metricId()

Returns this ErrorMetric’s unique ID string.

id

default_value

max_value

step

getValueStr(value)

Returns a human-readable string representation of the given value.

getNextValues(cur_value)

Returns a list of all possible values resulting from increasing cur_value by one step.

getPrevValues(cur_value)

Returns a list of all possible values from which cur_value can be reached in one step. The list is empty if this is not possible.

greaterMax(value)

Returns True if value exceeds max_value, else False.

convertToValue(value)

Converts the given string representation of a value into the correct value type.