Core Module Documentation¶

class neuralprophet.metrics.BatchMetric(name=None, specific_column=None)¶

Calculates a metric from batch model predictions.

abstract new(specific_column=None)¶

Parameters: specific_column (int) – Calculate metric only over target at pos
Returns: copy of metric instance
Return type: BatchMetric

update(predicted, target, **kwargs)¶

Updates the metric’s state using the passed batch output.

By default, this is called once for each batch.

Parameters

predicted (pd.DataFrame) – Output from the model’s forward function.
target (pd.DataFrame) – actual values
kwargs (dict) – Passed on to function that computes the metric.

class neuralprophet.metrics.LossMetric(loss_fn, specific_column=None)¶

Calculates the average loss according to the passed loss_fn.

Parameters: loss_fn (callable) – Taking a prediction tensor, a target tensor, optionally other arguments, and returns the average loss over all observations in the batch.

new(specific_column=None)¶

Parameters: specific_column (int) – Calculate metric only over target at pos
Returns: copy of metric instance
Return type: BatchMetric

class neuralprophet.metrics.MAE(specific_column=None, shift_scale=None)¶

Calculates the mean absolute error.

new(specific_column=None, shift_scale=None)¶

Parameters

specific_column (int) – Calculate metric only over target at pos
shift_scale (float, tuple) – Data shift and scale parameters

Returns

Copy of metric instance

Return type

MAE

set_shift_scale(shift_scale)¶

Adds data denormalization params

Parameters: shift_scale (float, tuple) – Data shift and scale parameters

class neuralprophet.metrics.MSE(specific_column=None, shift_scale=None)¶

Calculates the mean squared error.

new(specific_column=None, shift_scale=None)¶

Parameters

specific_column (int) – Calculate metric only over target at pos
shift_scale (float, tuple) – Data shift and scale parameters

Returns

copy of metric instance

Return type

MSE

set_shift_scale(shift_scale)¶

Adds data denormalization params.

Parameters: shift_scale (float, tuple) – Data shift and scale parameters

class neuralprophet.metrics.Metric(name=None)¶

Base class for all Metrics.

compute(save=False)¶

Calculates the current value of the metric

Parameters: save (bool) – Whether to add the current value to stored_values
Returns: Current value of the metric
Return type: float

print_stored()¶: Nice-prints stored values

reset(hard=False)¶

Resets the metric to it’s initial state.

By default, this is called at the start of each epoch.

set_shift_scale(shift_scale)¶: placeholder for subclasses to implement if applicable

abstract update(predicted, target)¶

Updates the metric’s state using the passed batch output.

By default, this is called once for each batch.

Parameters

predicted (pd.DataFrame) – Output from the model’s forward function.
target (pd.DataFrame) – Actual values

class neuralprophet.metrics.MetricsCollection(metrics, value_metrics=None)¶

Collection of Metrics that performs action over all

add_specific_target(target_pos)¶

Duplicates BatchMetrics with their version for a specific target.

Parameters: target_pos (int, list) – Index of target to compute metrics over

compute(save=False)¶

calculates the current value of the metric

Parameters: save (bool) – Whether to add the current value to stored_values
Returns: Current values of all metrics
Return type: dict

get_stored(loc=None)¶

Creates an OrderedDict from stored metric values

Parameters: loc (int) – If only stored value at this location to be retrieved
Returns: Current values of stored metric values
Return type: OrderedDict

get_stored_as_df(loc=None)¶

Creates an Dataframe from stored metric values

Parameters: loc (int) – If only stored value at this location to be retrieved
Returns: Current values of stored metric vlaues
Return type: pd.Dataframe

print(loc=None)¶: Nice-prints stored values

reset(hard=False)¶: reset all

set_shift_scale(shift_scale)¶

Adds data denormalization params to applicable metrics

Parameters: shift_scale (float, tuple) – Data shift and scale parameters

update(predicted, target, values=None)¶

Update all metrics.

Parameters

predicted (pd.DataFrame) – Output from the model’s forward function.
target (pd.DataFrame) – actual values
values (dict) –
Matching names to defined ValueMetrics

Note

If the correct name is not supplied, the metric is not updated.

update_batch(predicted, target)¶: update BatchMetrics

update_values(values, num)¶

Update ValueMetrics.

Parameters

values (dict) – containing matching names and average values over batch/update step
num (int) – number of samples in batch/update step

class neuralprophet.metrics.RMSE(specific_column=None, shift_scale=None)¶

Calculates the root mean squared error.

new(specific_column=None, shift_scale=None)¶

Parameters

specific_column (int) – Calculate metric only over target at pos
shift_scale (float) – Data shift and scale parameters

Returns

copy of metric instance

Return type

RMSE

set_shift_scale(shift_scale)¶

Adds data denormalization params.

Parameters: shift_scale (float, tuple) – Data shift and scale parameters

class neuralprophet.metrics.ValueMetric(name)¶

Keeps track of a value as a metric.

update(avg_value, num)¶

Parameters

avg_value (float) – average value over batch/update step
num (int) – number of samples in batch/update step