ml_prepare_response_features_intercept(x = NULL, response, features, intercept, envir = parent.frame(), categorical.transformations = new.env(parent = emptyenv()))ml_prepare_features(x, features, envir = parent.frame())
responseis a formula, it is used in preference to other parameters to set the
interceptparameters (if available). Currently, only simple linear combinations of existing parameters is supposed; e.g.
response ~ feature1 + feature2 + .... The intercept term can be omitted by using
- 1in the model fit.
interceptbindings should be mutated. (Typically, the parent frame).
Pre-process / normalize the inputs typically passed to a Spark ML routine.
Pre-processing of these inputs typically involves:
responseis itself a formula describing the model to be fit, thereby extracting the names of the
featuresto be used,
Please take heed of the last point, as while this is useful in practice, the behavior will be very surprising if you are not expecting it.