Model Results

ModelResult

class nupic.frameworks.opf.opf_utils.ModelResult(predictionNumber=None, rawInput=None, sensorInput=None, inferences=None, metrics=None, predictedFieldIdx=None, predictedFieldName=None, classifierInput=None)

A structure that contains the input to a model and the resulting predictions as well as any related information related to the predictions.

All params below are accesses as properties of the ModelResult object.

Parameters:

• predictionNumber – (int) This should start at 0 and increase with each new ModelResult.
• rawInput – (object) The input record, as input by the user. This is a dictionary-like object which has attributes whose names are the same as the input field names
• sensorInput – (SensorInput) object that represents the input record
• inferences – (dict) Each key is a InferenceType constant which corresponds to the type of prediction being made. Each value is a an element that corresponds to the actual prediction by the model, including auxillary information.
• metrics – (MetricsIface) The metrics corresponding to the most-recent prediction/ground truth pair
• predictedFieldIdx – (int) predicted field index
• predictedFieldName – (string) predicted field name
• classifierInput – (ClassifierInput) input from classifier

InferenceShifter

class nupic.data.inference_shifter.InferenceShifter

Shifts time for ModelResult objects. This is useful for plotting results with the predictions at the same time step as the input data.

shift(modelResult)

Shift the model result and return the new instance.

Queues up the T(i+1) prediction value and emits a T(i) input/prediction pair, if possible. E.g., if the previous T(i-1) iteration was learn-only, then we would not have a T(i) prediction in our FIFO and would not be able to emit a meaningful input/prediction pair.

Parameters:modelResult – A ModelResult instance to shift. Returns:A ModelResult instance that has been shifted