deriveArray {crunch}

Make a Categorical Array or Multiple Response variable

Description

In most situations we recommend using deriveArray which leaves your subvariables in the dataset. makeArray removes component subvariables from your dataset. Array variables are composed of a set of "subvariables" bound together for display in the app. For example, you might have a set of survey questions that ask how the respondent would rate a TV show from 1-5. Array variables allow you to display all of their ratings in a compact table rather than a set of distinct variables.

Usage

deriveArray(subvariables, name, selections, numeric = NULL, ...)

makeArray(subvariables, name, ...)

makeMR(subvariables, name, selections, ...)

Arguments

subvariables

a list of Variable objects to bind together, or a Dataset subset which contains only the Variables to bind.

name

character, the name that the new Categorical Array variable should have.

selections

character (preferred, indicating the names of the categories), or numeric (indicating the IDs of the categories in the combined array, which may not be the same as in the original variables - also note that a category's ID is not the same thing as its numeric_value). Required for makeMR; optional for deriveArray; ignored in makeArray.

numeric

Logical indicating whether the array should be a numeric array or categorical array. NULL the default will guess numeric if all variables are known to be numeric and categorical if all are categorical. If any subvariables are created from expressions, then their type cannot be guessed and so numeric must be specified.

...

Optional additional attributes to set on the new variable.

Value

A VariableDefinition that when added to a Dataset will create the categorical-array or multiple-response variable. deriveArray will make a derived array expression (or a derived multiple response expression if selections are supplied), while makeArray and makeMR return an expression that "binds" variables together, removing them from independent existence.

Examples

## Not run: 
# Categorical Array - Variables from list of variables
ds$enjoy_cat2 <- deriveArray(
    list(ds$enjoy1, ds$enjoy2),
    "Enjoy activities"
)

# Categorical Array - Variables from var catalog
# (result is the same as `ds$enjoy_cat1` above)
ds$enjoy_cat2 <- deriveArray(
    ds[c("enjoy1", "enjoy2")],
    "Enjoy activities v2"
)

# Multiple Response (selections as character names)
ds$enjoy_mr1 <- deriveArray(
    list(ds$enjoy1, ds$enjoy2),
    "Enjoy activities very much or a little",
    selections = c("Very much", "A little")
)

# Numeric Array
ds$rating_numa <- deriveArray(
    list(ds$rating1, ds$rating2),
    "Activity Rating"
)

# Using VarDef to specify metadata (and thus needing to specify type)
ds$enjoy_mr <- deriveArray(
    list(
        VarDef(ds$enjoy1 == "Very much", name = "enjoy brand 1"),
        VarDef(ds$enjoy2 == "Very much", name = "enjoy brand 2")
    ),
    "Enjoy activities with custom names"
)

# Multiple Response (selections as ids, same as ds$enjoy_mr1)
# Be careful `ids(categories(ds$enjoy1))` is not necessarily the same as
# `values(categories(ds$enjoy1))`
ds$enjoy_mr1 <- deriveArray(
    list(ds$enjoy1, ds$enjoy2),
    "Enjoy activities very much or a little v2",
    selections = c(1, 2)
)

## End(Not run)