Benchmarking: Hadleyverse vs Julia

Extends base Julia by adding features needed for statistical computing

• Missing values
• Data arrays that efficiently store missing values
• Tabular dataset

http://dataframesjl.readthedocs.org/en/latest/introduction.html

Contains several common datasets that we are used to

using DataFrames, RDatasets
diamonds = dataset("ggplot2", "diamonds")

https://github.com/johnmyleswhite/RDatasets.jl

http://dataframesjl.readthedocs.org/en/latest/split_apply_combine.html

1. Summarising a dataset
2. Sampling from a dataset
3. Tidying a dataset

• In old-hadleyverse, accomplished with plyr::ddply and plyr::summarise
• In new-hadleyverse, accomplished with dplyr::summarise
• In Julia, accomplished with by()

• Not really an old hadleyverse… would just use base::sample()
• In new-hadleyverse, accomplished with dplyr::sample_n
• In Julia, accomplished with sample() as well

• In old-hadleyverse, accomplished with reshape2::melt
• In new-hadleyverse, accomplished with tidyr::gather
• In Julia, accomplished with stack()