computeAME: Compute Average Marginal Effects
In compstat-lmu/ame: Compute average marginal effects for all models

Computes the average marginal effects for specified features.

1 2	computeAME(model, data, features, at = NULL, predict.fun = NULL, cl = NULL, ...)

`model`	[`WrappedModel` \| `any`] A model object. Can be also an object of class `WrappedModel`.
`data`	[`logical(1)`] The data set that was used to fit the model.
`features`	[`logical(1)`] The features for which the average marginal effects should be computed.
`at`	[`list`] (optional) A named list of vectors where the values specify at which points the marginal effects are calculated (i.e. the values are held constant).
`predict.fun`	[`function(1)`] The function that should be used to generate predictions from `model`. The specified prediction function is subject to subsequent numeric differentiation and thus needs to be carefully chosen in order to receive the correct results. This function must have two arguments, `object` and `newdata`. The default is the predict method for `model`. If `model` is of class `WrappedModel`, the default tries to use `getPredictionProbabilities` or `getPredictionResponse` depending on whether `model` is a classification or regression problem.
`cl`	('character(1)') Names of classes. Default is either all classes for multi-class / multilabel problems or the positive class for binary classification.
`...`	Further options passed down to the `grad` function.