Skip to contents

Constructs a new multiverse object which enables conducting a multiverse analysis

Usage

multiverse()

is_multiverse(x)

is.multiverse(x)

Arguments

x

An object

Value

An empty multiverse object

`TRUE` if the object inherits from the `multiverse` class.

`TRUE` if the object inherits from the `multiverse` class.

Details

To perform a multiverse analysis, the user needs to first constructs a new multiverse object. The user can declare multiple analysis pathways within this multiverse object, and execute these analyses. The multiverse object contains the following slots:

  • code: This slot stores the user's code for conducting a multiverse analysis. The user can define multiple analysis pathways at each step in the analysis using the `branch` call. It supports tidyverse syntax.

  • parameters: This slot contains a named list of lists. It contains each parameter defined in the code using the `branch()` function, and the options defined for each parameter (as a list).

  • conditions: This slot contain a list of conditions: if any of the parameter values are conditional on a specific value of another parameter, these can be defined in the code using `branch_assert()` or `branch_exclude()`.

  • current_parameter_assignment: This slot is a list which contains a single option assigned for each parameter defined in the `code`.

  • multiverse_table: This slot contains a table (in implementation, a tibble which is a rectangular data structure) where each column of the table will be a unique parameter. The table will contains every possible combination of options for each parameter — the number of rows corresponds to the number of different analysis paths. The table also contains, for each row, a list of option assignments for each parameter (`parameter_assignment` column), code for executing that particular analysis (of type `expression`), and environments where each code will be executed.

Examples

M <- multiverse()