Core Module Documentation#

neuralprophet.plot_model_parameters.plot_custom_season(m, comp_name, quantile, ax=None, figsize=(10, 6), df_name='__df__')#

Plot any seasonal component of the forecast.

Parameters:
  • m (NeuralProphet) – Fitted model

  • comp_name (str) – Name of seasonality component

  • quantile (float) – The quantile for which the custom season is plotted

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • focus (int) – Show weights for this forecast, if provided

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

  • df_name (str) –

    Name of dataframe to refer to data params from original keys of train dataframes

    Note

    Only used for local normalization in global modeling

Returns:

List of Artist objects containing seasonal forecast component

Return type:

matplotlib.artist.Artist

neuralprophet.plot_model_parameters.plot_daily(m, quantile, comp_name='daily', quick=True, ax=None, figsize=(10, 6), df_name='__df__')#

Plot the daily component of the forecast.

Parameters:
  • m (NeuralProphet) – Fitted model

  • quantile (float) – The quantile for which the daily seasonality is plotted

  • comp_name (str) – Name of seasonality component if previously changed from default daily

  • quick (bool) – Use quick low-level call of model

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

  • df_name (str) –

    Name of dataframe to refer to data params from original keys of train dataframes

    Note

    Only used for local normalization in global modeling

Returns:

List of Artist objects containing weekly forecast component

Return type:

matplotlib.artist.Artist

neuralprophet.plot_model_parameters.plot_lagged_weights(weights, comp_name, focus=None, ax=None, figsize=(10, 6))#

Make a barplot of the importance of lagged inputs.

Parameters:
  • weights (list) – tuples of (name, weights)

  • comp_name (str) – Name of lagged inputs

  • focus (int) – Show weights for this forecast, if provided

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

Returns:

List of Artist objects containing barplot

Return type:

matplotlib.artist.Artist

neuralprophet.plot_model_parameters.plot_parameters(m, quantile, forecast_in_focus=None, weekly_start=0, yearly_start=0, figsize=None, df_name=None)#

Plot the parameters that the model is composed of, visually.

Parameters:
  • m (NeuralProphet) – Fitted model

  • quantile (float) – The quantile for which the model parameters are to be plotted

  • forecast_in_focus (int) – n-th step ahead forecast AR-coefficients to plot

  • weekly_start (int) –

    Specifying the start day of the weekly seasonality plot

    Options
    • (default) weekly_start = 0: starts the week on Sunday

    • weekly_start = 1: shifts by 1 day to Monday, and so on

  • yearly_start (int) –

    Specifying the start day of the yearly seasonality plot.

    Options
    • (default) yearly_start = 0: starts the year on Jan 1

    • yearly_start = 1: shifts by 1 day to Jan 2, and so on

  • figsize (tuple) –

    Width, height in inches.

    Note

    Default value is set to None -> automatic figsize = (10, 3 * npanel)

  • df_name (str) –

    Name of dataframe to refer to data params from original keys of train dataframes

    Note

    Only used for local normalization in global modeling

Returns:

Figure showing the NeuralProphet parameters

Return type:

matplotlib.pyplot.figure

Examples

Base usage of plot_parameters()

>>> from neuralprophet import NeuralProphet
>>> m = NeuralProphet()
>>> metrics = m.fit(df, freq="D")
>>> future = m.make_future_dataframe(df=df, periods=365)
>>> forecast = m.predict(df=future)
>>> fig_param = m.plot_parameters()
neuralprophet.plot_model_parameters.plot_scalar_weights(weights, plot_name, focus=None, ax=None, figsize=(10, 6))#

Make a barplot of the regressor weights.

Parameters:
  • weights (list) – tuples of (name, weights)

  • plot_name (str) – Name of the plot Title

  • focus (int) – Show weights for this forecast, if provided

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

Returns:

List of Artist objects containing barplot

Return type:

matplotlib.artist.Artist

neuralprophet.plot_model_parameters.plot_trend(m, quantile, ax=None, plot_name='Trend', figsize=(10, 6), df_name='__df__')#

Make a barplot of the magnitudes of trend-changes.

Parameters:
  • m (NeuralProphet) – Fitted model

  • quantile (float) – The quantile for which the trend changes are plotted

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • plot_name (str) – Name of the plot Title

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

  • df_name (str) –

    Name of dataframe to refer to data params from original keys of train dataframes

    Note

    Only used for local normalization in global modeling

Returns:

List of Artist objects containing barplot

Return type:

matplotlib.artist.Artist

neuralprophet.plot_model_parameters.plot_trend_change(m, quantile, ax=None, plot_name='Trend Change', figsize=(10, 6), df_name='__df__')#

Make a barplot of the magnitudes of trend-changes.

Parameters:
  • m (NeuralProphet) – Fitted model

  • quantile (float) – The quantile for which the trend changes are plotted

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • plot_name (str) – Name of the plot Title

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

  • df_name (str) –

    Name of dataframe to refer to data params from original keys of train dataframes

    Note

    Only used for local normalization in global modeling

Returns:

List of Artist objects containing barplot

Return type:

matplotlib.artist.Artist

neuralprophet.plot_model_parameters.plot_weekly(m, quantile, comp_name='weekly', weekly_start=0, quick=True, ax=None, figsize=(10, 6), df_name='__df__')#

Plot the weekly component of the forecast.

Parameters:
  • m (NeuralProphet) – Fitted model

  • quantile (float) – The quantile for which the weekly seasonality is plotted

  • comp_name (str) – Name of seasonality component

  • weekly_start (int) –

    Specifying the start day of the weekly seasonality plot

    Options
    • (default) weekly_start = 0: starts the week on Sunday

    • weekly_start = 1: shifts by 1 day to Monday, and so on

  • quick (bool) – Use quick low-level call of model

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

  • df_name (str) –

    Name of dataframe to refer to data params from original keys of train dataframes

    Note

    Only used for local normalization in global modeling

Returns:

List of Artist objects containing weekly forecast component

Return type:

matplotlib.artist.Artist

neuralprophet.plot_model_parameters.plot_yearly(m, quantile, comp_name='yearly', yearly_start=0, quick=True, ax=None, figsize=(10, 6), df_name='__df__')#

Plot the yearly component of the forecast.

Parameters:
  • m (NeuralProphet) – Fitted model

  • quantile (float) – The quantile for which the yearly seasonality is plotted

  • comp_name (str) – Name of seasonality component

  • yearly_start (int) –

    Specifying the start day of the yearly seasonality plot

    Options
    • (default) yearly_start = 0: starts the year on Jan 1

    • yearly_start = 1: shifts by 1 day to Jan 2, and so on

  • quick (bool) – Use quick low-level call of model

  • ax (matplotlib axis) – Matplotlib Axes to plot on

  • figsize (tuple) –

    Width, height in inches, ignored if ax is not None.

    Note

    Default value is set to figsize = (10, 6)

  • df_name (str) –

    Name of dataframe to refer to data params from original keys of train dataframes

    Note

    Only used for local normalization in global modeling

Returns:

List of Artist objects containing yearly forecast component

Return type:

matplotlib.artist.Artist