Generic method for converting objects to Crunch representations

Source: R/R-to-variable.R

R objects are converted to Crunch objects using the following rules:

toVariable(x, ...)

# S4 method for class 'CrunchVarOrExpr'
toVariable(x, ..., derived = derivedVariableDefault())

# S4 method for class 'character'
toVariable(x, ...)

# S4 method for class 'numeric'
toVariable(x, ...)

# S4 method for class 'factor'
toVariable(x, ...)

# S4 method for class 'Date'
toVariable(x, ...)

# S4 method for class 'POSIXt'
toVariable(x, ...)

# S4 method for class 'AsIs'
toVariable(x, ...)

# S4 method for class 'VariableDefinition'
toVariable(x, ...)

# S4 method for class 'logical'
toVariable(x, ...)

# S4 method for class 'labelled'
toVariable(x, ...)

# S4 method for class 'haven_labelled'
toVariable(x, ...)

# S4 method for class 'labelled_spss'
toVariable(x, ...)

# S4 method for class 'haven_labelled_spss'
toVariable(x, ...)

Arguments

x

An R vector you want to turn into a Crunch variable

...

Additional metadata fields for the variable, such as "name" and "description". See the API documentation for a complete list of valid attributes.

derived

Logical, when FALSE indicates a variable should be materialized on creation (saved as data, which can have performance benefits in certain situation) and when TRUE indicates it should remain derived (saved as an expression that can update along with the underlying data) Defaults to TRUE unless envOrOption('crunch.default.derived') has been set.

Value

A VariableDefinition object. To add this to a dataset, either assign it into the dataset (like ds$newvar <- toVariable(...)) or call addVariables(). If you're adding a column of data to a dataset, it must be as long as the number of rows in the dataset, or it may be a single value to be recycled for all rows.

Details

  • Character vectors are converted into Crunch text variables

  • Numeric vectors are converted into Crunch numeric variables

  • Factors are converted to categorical variables

  • Date and POSIXt vectors are converted into Crunch datetime variables

  • Logical vectors are converted to Crunch categorical variables

  • VariableDefinition()s are not converted, but the function can still append additional metadata

If you have other object types you wish to convert to Crunch variables, you can declare methods for toVariable.

Examples

var1 <- rnorm(10)
toVariable(var1)
#> $values
#>  [1]  1.1484116 -1.8218177 -0.2473253 -0.2441996 -0.2827054 -0.5536994
#>  [7]  0.6289820  2.0650249 -1.6309894  0.5124269
#> 
#> $type
#> [1] "numeric"
#> 
#> attr(,"class")
#> [1] "VariableDefinition"
toVariable(var1, name = "Random", description = "Generated in R")
#> $values
#>  [1]  1.1484116 -1.8218177 -0.2473253 -0.2441996 -0.2827054 -0.5536994
#>  [7]  0.6289820  2.0650249 -1.6309894  0.5124269
#> 
#> $type
#> [1] "numeric"
#> 
#> $name
#> [1] "Random"
#> 
#> $description
#> [1] "Generated in R"
#> 
#> attr(,"class")
#> [1] "VariableDefinition"
if (FALSE) { # \dontrun{
ds$random <- toVariable(var1, name = "Random")
# Or, this way:
ds <- addVariables(ds, toVariable(var1, name = "Random"))
} # }