Laurae2 R-package

The sequel to Laurae R-package.

Each function has at least one corresponding vignette to look up for an example using help_me("my_function_name").

Installation

It can be computationally expensive to build vignettes. Build without vignettes using the following:

devtools::install_github("Laurae2/Laurae2")

If you want to build vignettes to get a significantly better help:

devtools::install_github("Laurae2/Laurae2", build_vignettes = TRUE)

Pre-requirement installation:

install.packages("devtools")
install.packages(c("knitr", "rmarkdown", "mlrMBO", "lhs", "smoof", "ParamHelpers", "animation"))

xgboost installation, commit dmlc/xgboost@017acf5 seems best currently as it has gblinear improvements. Make sure to use the right compiler below:

devtools::install_github("Laurae2/xgbdl")

# gcc
xgbdl::xgb.dl(compiler = "gcc", commit = "017acf5", use_avx = FALSE, use_gpu = FALSE)

# Visual Studio 2015, use AVX if you wish to
xgbdl::xgb.dl(compiler = "Visual Studio 14 2015 Win64", commit = "017acf5", use_avx = FALSE, use_gpu = FALSE)

# Visual Studio 2017, use AVX if you wish to
xgbdl::xgb.dl(compiler = "Visual Studio 15 2017 Win64", commit = "017acf5", use_avx = FALSE, use_gpu = FALSE)

What can it do?

What can it do:

• Bayesian Optimization (time-limited, iteration-limited, initialization-limited)
• Create data.frame from [R,C] matrix-like format
• Create data.table from [R,C] matrix-like format

Package requirements:

• knitr
• rmarkdown
• mlrMBO
• lhs
• smoof
• ParamHelpers
• animation
• xgboost