Utilities for developing R software

The {oeli} package offers a collection of handy functions that I found useful while developing R packages. Perhaps you’ll find them helpful too!

Installation

The released package version can be installed from CRAN via:

install.packages("oeli")

Demos

The package includes helpers for various tasks and objects. Some demos are shown below. Click the headings for reference pages with documentation on all available helpers in each category.

Distributions

The package has density and sampling functions for distributions not in base R, such as Dirichlet, multivariate normal, truncated normal, and Wishart. For faster computation, an Rcpp implementation is also available.

ddirichlet(x = c(0.2, 0.3, 0.5), concentration = 1:3)
#> [1] 4.5

rdirichlet(concentration = 1:3)
#> [1] 0.2032900 0.2310928 0.5656171

Function helpers

Retrieving default arguments of a function:

f <- function(a, b = 1, c = "", ...) { }
function_defaults(f)
#> $b
#> [1] 1
#> 
#> $c
#> [1] ""

Indexing helpers

Create all possible permutations of vector elements:

permutations(LETTERS[1:3])
#> [[1]]
#> [1] "A" "B" "C"
#> 
#> [[2]]
#> [1] "A" "C" "B"
#> 
#> [[3]]
#> [1] "B" "A" "C"
#> 
#> [[4]]
#> [1] "B" "C" "A"
#> 
#> [[5]]
#> [1] "C" "A" "B"
#> 
#> [[6]]
#> [1] "C" "B" "A"

Package helpers

Quickly have a basic logo for your new package:

package_logo("my_package", brackets = TRUE, use_logo = FALSE)

How to print a matrix without filling up the entire console?

x <- matrix(rnorm(10000), ncol = 100, nrow = 100)
print_matrix(x, rowdots = 4, coldots = 4, digits = 2, label = "what a big matrix")
#> what a big matrix : 100 x 100 matrix of doubles 
#>         [,1]  [,2]  [,3] ... [,100]
#> [1,]    0.01  0.59 -1.02 ...   1.73
#> [2,]   -0.11 -0.01  2.37 ...   0.22
#> [3,]   -1.87  0.09 -1.24 ...  -0.93
#> ...      ...   ...   ... ...    ...
#> [100,]   0.6 -0.99 -0.26 ...   -0.9

Simulation helpers

Let’s simulate a Markov chain:

Gamma <- sample_transition_probability_matrix(dim = 3)
simulate_markov_chain(Gamma = Gamma, T = 20)
#>  [1] 2 2 3 2 1 2 2 2 2 1 2 2 1 1 2 3 3 1 1 1

Transformation helpers

The group_data_frame() function groups a given data.frame based on the values in a specified column:

df <- data.frame("label" = c("A", "B"), "number" = 1:10)
group_data_frame(df = df, by = "label")
#> $A
#>   label number
#> 1     A      1
#> 3     A      3
#> 5     A      5
#> 7     A      7
#> 9     A      9
#> 
#> $B
#>    label number
#> 2      B      2
#> 4      B      4
#> 6      B      6
#> 8      B      8
#> 10     B     10

Validation helpers

Is my matrix a proper transition probability matrix?

matrix <- diag(4)
matrix[1, 2] <- 1
check_transition_probability_matrix(matrix)
#> [1] "Must have row sums equal to 1"