How to convert categorical string data to numeric data

Sometimes, categorical data are stored as strings. For example, the variable race may be stored with the words "Black", "Other", and "White". We will need to convert these variables to numeric data before we can use them with Stata's statistical features.

Let's begin by opening and describing an example dataset from the Stata website.

. use https://www.stata.com/users/youtube/rawdata.dta, clear
(Fictitious data based on the National Health and Nutrition Examination Survey)

. describe

Contains data from https://www.stata.com/users/youtube/rawdata.dta
 Observations:         1,268                  Fictitious data based on the
                                                National Health and Nutrition
                                                Examination Survey
    Variables:            10                  6 Jul 2016 11:17
                                              (_dta has notes)

 
Variable      Storage   Display    Value
    name         type    format    label      Variable label
 
id              str6    %9s                   Identification Number
age             byte    %9.0g                 
sex             byte    %9.0g                 Sex
race            str5    %9s                   Race
height          float   %9.0g                 height (cm)
weight          float   %9.0g                 weight (kg)
sbp             int     %9.0g                 Systolic blood pressure (mm/Hg)
dbp             int     %9.0g                 Diastolic blood pressure (mm/Hg)
chol            str3    %9s                   serum cholesterol (mg/dL)
dob             str18   %18s                  
 
Sorted by: id

The storage type for the variable race is a 5-character string. Let's tabulate race to view the categories.

. tabulate race


       Race 
 
      Freq.     Percent        Cum.
 
 
 
      Black 
 
        176       13.88       13.88
      Other 
 
         22        1.74       15.62
      White 
 
      1,070       84.38      100.00
 
 
 
      Total 
 
      1,268      100.00            

There are three categories stored as the strings: Black, Other, and White. We can use Stata's encode command to generate a new variable named racen.

. encode race, gen(racen)

Let's type tabulate race racen to view a cross-tabulation of the two variables and list race racen in 1/5 to view some raw data.

. tabulate race racen

. list race racen in 1/5


     
  race   racen 
  1. 
 White   White 
  2. 
 White   White 
  3. 
 White   White 
  4. 
 Black   Black 
  5. 
 White   White 

The two variables appear to be identical. Next let's describe both variables.

. describe race racen


Variable      Storage   Display    Value
    name         type    format    label      Variable label
 
race            str5    %9s                   Race
racen           long    %8.0g      racen      Race

The storage type for race is "str5", and the storage type for racen is "long", which is a type of numeric variable. You can type help data_types to learn more about different types of numeric data. Notice that the Value label for racen is "racen". Let's type label list racen to view the labels.

. label list racen
racen:
           1 Black
           2 Other
           3 White

The variable racen is a numeric variable where 1 represents Black, 2 represents Other, and 3 represents White. This will allow us to use racen with Stata's statistical features such as regression modeling.

Note that there is a decode command that will do the reverse of encode: it will convert labeled numeric categorical variables to string variables.

. decode racen, gen(races)

We can use describe and list to verify that it worked.

. describe race racen races

. list race racen races in 1/5


     
  race   racen   races
  1. 
 White   White   White
  2. 
 White   White   White 
  3. 
 White   White   White 
  4. 
 Black   Black   Black 
  5. 
 White   White   White

The raw data look the same for all three variables, but, as we have learned, the storage type is important. And now we know how to convert between types when necessary.

You can watch a demonstration of these commands by clicking on the link to the YouTube video below. You can read more about these commands by clicking on the links to the Stata manual entries below.

See it in action

Watch Data management: How to convert categorical string variables to labeled numeric variables.