Learn R Programming

modelenv

modelenv is a developer focused package designed to provide tools to register models much in the same as is done in parsnip. This package does not replace parsnip’s registration functions, but will be used in tidyclust and future non-parsnip modeling packages to avoid complex dependencies.

Installation

You can install the released version of modelenv from CRAN with:

install.packages("modelenv")

And the development version from GitHub with:

# install.packages("pak")
pak::pak("tidymodels/modelenv")

Learn more

These functions have near identical interfaces to their parsnip equivalent, see tidymodels.org - How to build a parsnip model for instructions of how to use these functions.

Copy Link

Version

Install

install.packages('modelenv')

Monthly Downloads

28,649

Version

0.2.0

License

MIT + file LICENSE

Issues

Pull Requests

Stars

Forks

Maintainer

Emil Hvitfeldt

Last Published

October 14th, 2024

Functions in modelenv (0.2.0)

stop_incompatible_mode

Error handling for incompatible modes
get_model_env

Working with the modelenv model environment
modelenv-package

modelenv: Provide Tools to Register Models for Use in 'tidymodels'
set_dependency

Register Dependency for Model
new_unsupervised_spec

Give object unsupervised specification class
check_spec_mode_engine_val

Error handling for unknown mode
is_unsupervised_fit

Determine object is has class unsupervised_fit
set_fit

Register Fit method for Model
new_unsupervised_fit

Give object unsupervised fit class
is_unsupervised_spec

Determine object is has class unsupervised_spec
set_encoding

Register Encoding Options for Model
set_pred

Register Prediction Method for Model
set_model_mode

Register Mode for Model
set_new_model

Register New Model
set_model_engine

Register Engine for Model
set_model_arg

Register Argument for Model