Convert Multiple Columns to Attribute Combinations

Creates combined attribute values from multiple columns in a synthetic population or contingency table.

Usage

multicolumn_to_attribute_values(df, attr_name, columns)

Arguments

df

Input data frame or data.table.

attr_name

Character string specifying the name of the generated attribute.

columns

Character vector containing the columns to convert.

Value

A character vector containing combined attribute values.

Details

This function is useful when constructing composite grouping variables from several demographic attributes.

Examples include:

• Age-group and gender combinations.

• Education and employment combinations.

• Geographic and demographic combinations.

Values from the supplied columns are combined into a single string representation for each row.

The resulting values can be used as:

• Composite identifiers.

• Grouping variables.

• Keys for matching and aggregation.

This function is used internally by several utilities in replica but may also be useful for user-defined reporting and validation workflows.

See also

synthetic_population_to_contingency, get_margin_series_from_synthetic_population

Examples

df <- data.frame(
  Degree = c(50, 20),
  Diploma = c(30, 10),
  School = c(20, 70),
  gender = c(
    "Male",
    "Female"
  )
)

multicolumn_to_attribute_values(
  df = df,
  attr_name = "education",
  columns = c(
    "Degree",
    "Diploma",
    "School"
  )
)
#>    gender education count
#>    <char>    <fctr> <num>
#> 1:   Male    Degree    50
#> 2: Female    Degree    20
#> 3:   Male   Diploma    30
#> 4: Female   Diploma    10
#> 5:   Male    School    20
#> 6: Female    School    70