Learn R Programming

⚠️There's a newer version (0.3.9) of this package.Take me there.

manymodelr (version 0.3.2)

Build and Tune Several Models

Description

Frequently one needs a convenient way to build and tune several models in one go.The goal is to provide a number of machine learning convenience functions. It provides the ability to build, tune and obtain predictions of several models in one function. The models are built using 'caret' functions with easier to read syntax. Kuhn(2014) .

Copy Link

Version

Install

install.packages('manymodelr')

Monthly Downloads

551

Version

0.3.2

License

GPL-2

Issues

Pull Requests

Stars

Forks

Maintainer

Nelson Gonzabato

Last Published

June 27th, 2020

Functions in manymodelr (0.3.2)

add_model_predictions

Add predictions to the data set. A dplyr compatible way to add predictions to a data set.
add_model_residuals

Add model residuals
na_replace

Replace missing values
get_mode

A convenience function that returns the mode
extract_model_info

Extract important model attributes
agg_by_group

A convenient way to perform grouped operations
multi_model_1

Simultaneously train and predict on new data.
get_data_Stats

A pipe friendly way to get summary stats for exploratory data analysis
multi_model_2

Fit and predict in one function
get_this

Helper function to easily access elements
get_var_corr

Get correlations between variables
na_replace_grouped

Replace NAs by group
get_exponent

Get the exponent of any number or numbers
fit_models

Fit several models with different response variables
fit_model

Fit and predict in a single function.
select_col

A convenient selector gadget
plot_corr

Plot a correlations matrix
get_var_corr_

Get correlations for combinations
rowdiff

Get row differences between values
select_percentile

Get the row corresponding to a given percentile