categorize

Data exploration and transformation

Create a categorical variable from a continuous one.

Description

categorize is a shortcut and extension of egen newvar = cut(args) [...], icodes.

Unlike egen with the cut() function, categorize:

Does not require the user to include the minimum and the maximum value of the continuous variable in the list of breaks.
Creates more descriptive value labels for the generated categorical variable. Users can specify a variable label for the new variable.
Allows users working with age or poverty ratio variables to use “default” breaks.

categorize varname, generate(newvar) {breaks(numlist)|default(string)} [options]

Option	Description
`generate(newvar)`	Name of the categorical variable to be generated.
`breaks(numlist)`	Left-hand ends of the grouping intervals. Do not include the minimum or the maximum value of varname. Either `breaks()` or `default()` must be specified.
`default(string)`	Use default breaks; “age” or “povratio”. For `default("age")`, these are 18 and 65. For `default("povratio")`, these are 50, 100, 150, 200, and 250. Cannot be combined with `breaks()`.
`lblname(string)`	Name of value label to create; default is “varname_lbl”. Ignored if `nolabel` is specified.
`nformat(%fmt)`	Numeric format to use in value labels; default is %13.0gc. Ignored if `nolabel` is specified.
`nolabel`	Do not assign value labels to newvar.
`varlabel(string)`	Variable label for newvar.

Using user-specified breaks.

categorize pincp_adj, generate(pincp_cat) breaks(25000 50000 100000)

Using default breaks.

categorize agep, generate(age_cat) default("age") varlabel("Age group")