====== Problem ======

Some R packages provide generic functions e.g. for training models, calculating distances, etc., where one model out of a set of various possible models has to be chosen as a parameter of the function.

Issues:
  * Documenting the possible entries is not easy, because a single model description already contains a lot of information. When many models must be described in help, the help page of the function becomes too large (in particular the parameter description part).
  * The model description itself should be structured.

====== Existing solutions ======

How is this problem solved by different popular packages?

===== caret =====
==== train method ====
  method: a string specifying which classification or regression model
          to use. Possible values are found using
          ‘names(getModelInfo())’. See <URL:
          http://topepo.github.io/caret/bytag.html>. A list of
          functions can also be passed for a custom model function. See
          <URL: http://topepo.github.io/caret/custom_models.html> for
          details.


  preProcess: a string vector that defines a pre-processing of the
          predictor data. Current possibilities are "BoxCox",
          "YeoJohnson", "expoTrans", "center", "scale", "range",
          "knnImpute", "bagImpute", "medianImpute", "pca", "ica" and
          "spatialSign". The default is no pre-processing. See
          ‘preProcess’ and ‘trainControl’ on the procedures and how to
          adjust them. Pre-processing code is only designed to work
          when ‘x’ is a simple matrix or data frame.


getModelInfo shows infos about models and packages that are accessible via ‘train’

  Usage:
     modelLookup(model = NULL)
     
     checkInstall(pkg)
     
     getModelInfo(model = NULL, regex = TRUE, ...)
     
     ‘modelLookup’ is good for getting information related to the
     tuning parameters for a model. ‘getModelInfo’ will return all the
     functions and metadata associated with a model. Both of these
     functions will only search within the models bundled in this
     package.

  Value:
   ‘modelLookup’ produces a data frame with columns
   
   model: a character string for the model code
   
   parameter : the tuning parameter name
   
   label : a tuning parameter label (used in plots)
   
   forReg : a logical; can the model be used for regression?
   
   forClass : a logical; can the model be used for classification?
   
   probModel : a logical; does the model produce class probabilities?
     ‘getModelInfo’ returns a list containing one or more lists of the
     standard model information.
   
Returned info from getModelInfo is a rather cryptic list.

preProcess is a function which has it's own different method types. The methods are explained in the details section of the perProcess help page.

trainControl is also a function, where different possible values for methods are listed, but not explained. Only hints which methods should be used under which circumstances.

===== proxy =====

  # summary of available distance measures
  summary(pr_DB)
  # particular info about one distance measure
  pr_DB$get_entry("Jaccard")
  
===== dtw =====
  # directly print stepPattern (it's a class)
  print(symmetric2)
  
===== arules =====
  object of class APparameter or named list.