Skip to contents

vazul

vazul is an R package that provides functions for data blinding, allowing researchers to scramble data while preserving its statistical properties.

Installation

You can install the development version of vazul from GitHub with:

# install.packages("devtools")
devtools::install_github("nthun/vazul")

Main Functions

scramble_values()

Scrambles the values within a vector without replacement, preserving the original values but changing their order.

library(vazul)

# Scramble numeric values
set.seed(123)
scramble_values(1:10)
#>  [1]  3 10  2  8  6  9  1  7  5  4

# Scramble character values  
scramble_values(letters[1:5])
#> [1] "b" "c" "a" "d" "e"

# Scramble logical values
scramble_values(c(TRUE, FALSE, TRUE, FALSE))
#> [1] FALSE  TRUE FALSE  TRUE

scramble_variables()

Scrambles the values of specified variables (columns) within a data frame, with optional grouping.

# Create example data
df <- data.frame(
  x = 1:6, 
  y = letters[1:6], 
  group = c("A", "A", "A", "B", "B", "B")
)

# Scramble variables across entire dataset
scramble_variables(df, c("x", "y"))

# Scramble within groups using .groups parameter
scramble_variables(df, "y", .groups = "group")

# Or using dplyr grouping
library(dplyr)
df |> 
  group_by(group) |> 
  scramble_variables("x") |>
  ungroup()
  
# Can use multiple groups
marp |> 
    group_by(country, gender) |> 
    scramble_variables(c("rel_1", "rel_2")) |> 
    unroup()

Included Datasets

MARP Dataset

The Many Analysts Religion Project (MARP) dataset contains data from 10,535 participants across 24 countries, exploring the relationship between religiosity and well-being.

data(marp)
dim(marp)
#> [1] 10535    46

# Explore countries in the dataset
length(unique(marp$country))
#> [1] 24

# Example: scramble religiosity variables by country
marp |>
  group_by(country) |>
  scramble_variables(c("rel_1", "rel_2", "rel_3")) |>
  ungroup()

Variables include: - rel_1 to rel_9: Religiosity measures - wb_*: Well-being indicators (general, physical, psychological, social) - country: Participant country - age, gender, ses, education: Demographics - gdp: Country-level GDP data

Williams Dataset

A study dataset with 112 participants examining the risk taking behavior behavior of high or low wealth individuals.

data(williams)
dim(williams)
#> [1] 112  25

table(williams$ecology)
#> Desperate   Hopeful 
#>        56        56

# Example: scramble perception measures within ecology conditions
williams |>
  group_by(ecology) |>
  scramble_variables(c("SexUnres_1", "Impuls_1")) |>
  ungroup()

Key variables: - ecology: Experimental condition (“Desperate” vs “Hopeful”) - SexUnres_*: Sexual unresponsiveness measures
- Impuls_*: Impulsivity measures - Opport_*: Long-term planning opportunity measures - age, gender: Participant demographics

Explanation of the package name

Vazul was a historical figure (Hungarian price) in the 11. century. He was blinded by the king to become unfit for the throne. More info: https://en.wikipedia.org/wiki/Vazul

Documentation

Authors

  • Tamás Nagy - Package author and maintainer
  • Alexandra Sarafoglou - Data contributor and author
  • Márton Kovács - Author

License

This project is licensed under the MIT License - see the LICENSE file for details.