README.md

beetroot

Travis build status Codecov test coverage Lifecycle: experimental

The goal of beetroot is to enable the parsnip package to fit mixture models.

Installation

You can install the development version from GitHub with:

# install.packages("devtools")
devtools::install_github("hfrick/beetroot")

Example

This is a basic example of fitting a mixture model and getting predictions:

library(beetroot)
#> Loading required package: parsnip
library(rsample)
#> Loading required package: tidyr

set.seed(403)
data("faithful")
faithful_split <- initial_split(faithful, prop = 0.95)
faithful_train <- training(faithful_split)
faithful_test <- testing(faithful_split)

mm_spec <- mixture_model(components = 2) %>% 
  set_engine("flexmix")

mm_fit <- mm_spec %>%
  fit(eruptions ~ waiting, data = faithful_train)

predict(mm_fit, new_data = faithful_test, type = "numeric") 
#> # A tibble: 13 x 1
#>    .pred
#>    <dbl>
#>  1  4.03
#>  2  1.68
#>  3  1.76
#>  4  2.67
#>  5  3.50
#>  6  3.88
#>  7  2.67
#>  8  4.33
#>  9  3.96
#> 10  2.44
#> 11  5.02
#> 12  4.64
#> 13  2.21


hfrick/beetroot documentation built on April 13, 2020, 12:39 a.m.