intubate (version 1.0.0)

glmnet: Interfaces for glmnet package for data science pipelines.

Description

Interfaces to glmnet functions that can be used in a pipeline implemented by magrittr.

Usage

ntbt_glmnet(data, ...)
ntbt_cv.glmnet(data, ...)

Arguments

data

data frame, tibble, list, ...

...

Other arguments passed to the corresponding interfaced function.

Value

Object returned by interfaced function.

Details

Interfaces call their corresponding interfaced function.

Examples

## Not run: 
# library(intubate)
# library(magrittr)
# library(glmnet)
# 
# 
# ## NOTE: glmnet package does not implement formula interface. As I need it
# ##       for teaching purposes, this is the first pure non-formula library
# ##       included, as a proof of concept that not only the functions with
# ##       formula + data variant can be successfully interfaced to use in
# ##       a pipeline.
# 
# 
# library(ISLR)
# data("Hitters")
# Hitters <- na.omit(Hitters)
# 
# dta <- list(x = model.matrix(Salary ~ ., Hitters)[, -1],  ## Remove intercept
#             y = model.frame(Salary ~ ., Hitters)[, 1])
# grid <- 10^seq(10, -2, length = 100)
# 
# ## ntbt_glmnet: fit a GLM with lasso or elasticnet regularization
# 
# ## Original function to interface
# attach(dta)
# ## Ridge Regression
# ridge <- glmnet(x, y, alpha = 0, lambda = grid)
# plot(ridge)
# 
# ## The Lasso
# lasso <- glmnet(x, y, alpha = 1, lambda = grid)
# plot(lasso)
# detach()
# 
# ## The interface puts data as first parameter
# ## Ridge Regression
# ridge <- ntbt_glmnet(dta, x, y, alpha = 0, lambda = grid)
# plot(ridge)
# 
# ## The Lasso
# lasso <- ntbt_glmnet(dta, x, y, alpha = 1, lambda = grid)
# plot(lasso)
# 
# ## so it can be used easily in a pipeline.
# ## Ridge Regression
# dta %>%
#   ntbt_glmnet(x, y, alpha = 0, lambda = grid) %>%
#   plot()
# 
# ## The Lasso
# dta %>%
#   ntbt_glmnet(x, y, alpha = 1, lambda = grid) %>%
#   plot()
# 
# 
# ## ntbt_cv.glmnet: Cross-validation for glmnet
# 
# ## Original function to interface
# attach(dta)
# ## Ridge Regression
# set.seed(1)
# cv.ridge <- cv.glmnet(x, y, alpha = 0)
# plot(cv.ridge)
# 
# ## The Lasso
# cv.lasso <- cv.glmnet(x, y, alpha = 1)
# plot(cv.lasso)
# detach()
# 
# ## The interface puts data as first parameter
# ## Ridge Regression
# set.seed(1)
# cv.ridge <- ntbt_cv.glmnet(dta, x, y, alpha = 0)
# plot(cv.ridge)
# 
# ## The Lasso
# cv.lasso <- ntbt_cv.glmnet(dta, x, y, alpha = 1)
# plot(cv.lasso)
# 
# ## so it can be used easily in a pipeline.
# ## Ridge Regression
# set.seed(1)
# dta %>%
#   ntbt_cv.glmnet(x, y, alpha = 0) %>%
#   plot()
# 
# ## The Lasso
# dta %>%
#   ntbt_cv.glmnet(x, y, alpha = 1) %>%
#   plot()
# ## End(Not run)