Skip to content

Model Specifications

Source: R/model_object_docs.R
model_spec.Rd

The parsnip package splits the process of fitting models into two steps:

  1. Specify how a model will be fit using a model specification

  2. Fit a model using the model specification

This is a different approach to many other model interfaces in R, like lm(), where both the specification of the model and the fitting happens in one function call. Splitting the process into two steps allows users to iteratively define model specifications throughout the model development process.

This intermediate object that defines how the model will be fit is called a model specification and has class model_spec. Model type functions, like linear_reg() or boost_tree(), return model_spec objects.

Fitted model objects, resulting from passing a model_spec to fit() or fit_xy, have class model_fit, and contain the original model_spec objects inside them. See ?model_fit for more on that object type, and ?extract_spec_parsnip to extract model_specs from model_fits.

Details

An object with class "model_spec" is a container for information about a model that will be fit.

The main elements of the object are:

  • args: A vector of the main arguments for the model. The names of these arguments may be different from their counterparts n the underlying model function. For example, for a glmnet model, the argument name for the amount of the penalty is called "penalty" instead of "lambda" to make it more general and usable across different types of models (and to not be specific to a particular model function). The elements of args can tune() with the use of the tune package. For more information see https://www.tidymodels.org/start/tuning/. If left to their defaults (NULL), the arguments will use the underlying model functions default value. As discussed below, the arguments in args are captured as quosures and are not immediately executed.

  • ...: Optional model-function-specific parameters. As with args, these will be quosures and can be tune().

  • mode: The type of model, such as "regression" or "classification". Other modes will be added once the package adds more functionality.

  • method: This is a slot that is filled in later by the model's constructor function. It generally contains lists of information that are used to create the fit and prediction code as well as required packages and similar data.

  • engine: This character string declares exactly what software will be used. It can be a package name or a technology type.

This class and structure is the basis for how parsnip stores model objects prior to seeing the data.

Argument Details

An important detail to understand when creating model specifications is that they are intended to be functionally independent of the data. While it is true that some tuning parameters are data dependent, the model specification does not interact with the data at all.

For example, most R functions immediately evaluate their arguments. For example, when calling mean(dat_vec), the object dat_vec is immediately evaluated inside of the function.

parsnip model functions do not do this. For example, using


 rand_forest(mtry = ncol(mtcars) - 1)

does not execute ncol(mtcars) - 1 when creating the specification. This can be seen in the output:


 > rand_forest(mtry = ncol(mtcars) - 1)
 Random Forest Model Specification (unknown)

 Main Arguments:
   mtry = ncol(mtcars) - 1

The model functions save the argument expressions and their associated environments (a.k.a. a quosure) to be evaluated later when either fit.model_spec() or fit_xy.model_spec() are called with the actual data.

The consequence of this strategy is that any data required to get the parameter values must be available when the model is fit. The two main ways that this can fail is if:

  1. The data have been modified between the creation of the model specification and when the model fit function is invoked.

  2. If the model specification is saved and loaded into a new session where those same data objects do not exist.

The best way to avoid these issues is to not reference any data objects in the global environment but to use data descriptors such as .cols(). Another way of writing the previous specification is


 rand_forest(mtry = .cols() - 1)

This is not dependent on any specific data object and is evaluated immediately before the model fitting process begins.

One less advantageous approach to solving this issue is to use quasiquotation. This would insert the actual R object into the model specification and might be the best idea when the data object is small. For example, using


 rand_forest(mtry = ncol(!!mtcars) - 1)

would work (and be reproducible between sessions) but embeds the entire mtcars data set into the mtry expression:


 > rand_forest(mtry = ncol(!!mtcars) - 1)
 Random Forest Model Specification (unknown)

 Main Arguments:
   mtry = ncol(structure(list(Sepal.Length = c(5.1, 4.9, 4.7, 4.6, 5, <snip>

However, if there were an object with the number of columns in it, this wouldn't be too bad:


 > mtry_val <- ncol(mtcars) - 1
 > mtry_val
 [1] 10
 > rand_forest(mtry = !!mtry_val)
 Random Forest Model Specification (unknown)

 Main Arguments:
   mtry = 10

More information on quosures and quasiquotation can be found at https://adv-r.hadley.nz/quasiquotation.html.