fredplot

fredplot(
    model,
    x,
    type='kde',
    scale='uniform',
    jitter=False,
    lowess=False,
    frac=2 / 3,
    n=None,
    ax=None,
    **kwargs,
)

Create a Functional REsidual Density (FRED) plot.

This plot shows the density of the functional residuals against a continuous predictor variable, which is useful for identifying model misspecifications like missing non-linear terms.

Parameters

Name	Type	Description	Default
model	object	A fitted model object from statsmodels.	required
x	np.ndarray or pd.Series	The predictor variable to plot on the x-axis.	required
type	str	The type of density plot to generate. One of {“kde”, “hex”}, by default “kde”.	'kde'
scale	str	The scale to use for the functional residuals. One of {“uniform”, “normal”}, by default “uniform”.	'uniform'
jitter	bool	If True, add a small amount of random noise to the predictor x. This is useful for visualizing categorical predictors. By default False.	False
lowess	bool	If True, add a LOWESS smooth to the plot, by default False.	False
frac	float	The fraction of data used when estimating each y-value of the LOWESS smooth, by default 2/3.	2 / 3
n	int	The number of observations to subsample, by default None (use all).	None
ax	matplotlib.axes.Axes	An existing matplotlib Axes object to plot on. If None, a new figure and axes are created.	None
**kwargs	dict	Additional keyword arguments passed to the seaborn plotting function (e.g., sns.kdeplot or ax.hexbin).	{}

Returns

Name	Type	Description
	matplotlib.axes.Axes	The matplotlib Axes object containing the plot.