Rust

Rust provides Ruby Statistic facilities. It mostly relies on R (rinruby gem) and it provides the most useful statistical functions in a Ruby wrapper.

To install the latest version of Rust, simply run the following command

gem install rust

Everything in Rust is defined in the Rust module and organized in submodules, one for each core feature. The modules currently available are:

• Rust contains the core data structures (data frames, matrices) and functions that allow to interact with R. The data structures provide their core functions (e.g., data frames have the merge method);
• Rust::Descriptive provides descriptive statistic functions (e.g., mean, variance);
• Rust::Correlation provides object to compute correlations (Pearson, Spearman, and Kendall);
• Rust::StatisticalTest allows to run statistical hypothesis tests (t-test, Wilcoxon);
• Rust::EffectSize allows to compute the effect size (Cohen d, Cliff's delta);
• Rust::CSV allows to read and write CSV files;
• Rust::Plots provides basic plotting features.

Each module can optionally define some R binding methods: such methods allow to provide an interface very similar to R. To use these bindings, just call the method bind_r!.

Here we provide some examples to quick-start using Rust.

# Load a data frame with columns id, name, surname, height
people = Rust::CSV.read("people.csv", headers: true)
# Sum the values of the column "test" of dataframe
p people.column("height").sum

# Load a data frame with columns id, amount
purchases = Rust::CSV.read("purchases.csv", headers: true)
# Merge the two dataframes in one with columns id, name, surname, height, purchase.amount
merged = people.merge(purchases, ["id"], "", "purchase")
# Gets the surnames of the the people who spent more than 100
p merged.select_rows { |r| r['purchase.amount'] > 100 }.column("surname") 

# Create a scatter plot object (does not show anything) with the x and y values
sp = Rust::Plots::ScatterPlot.new([1,2,3,4,5], [2,3,4,5,6])
# Will connect the points with the lines
sp.lines
# Show the plot
sp.show

# Create a box plot object (does not show anything)
bp = Rust::Plots::BoxPlot.new
# Add two series of data, one named 'hello', one named 'world'
bp.series([1,2,3,5,4,6], name: 'hello')
bp.series([2,4,6,3,5,4], name: 'world')
# Save the plot on the PDF file "result.pdf" with a with of 6 inches
bp.pdf("result.pdf", width: 6)