Creating a new categorical variable for which variables 1 and 2 contain text

[advmsj]

Hi there, I am trying to find a way to create a categorical variable in which:

1 = Mixed (given there is text in both columns 1 and 2)

2 = Unknown

3 = Purebred type 1

.

.

.

n = Purebred type n

Given the sample dataset:

ID	Column 1 - Breed 1	Column 2 - Breed 2	Desired Column
1	Poodle		Purebred - Poodle
2	Poodle	Husky	Mixed
3	Husky		Purebred - Husky
4	Maltese		Purebred - Maltese
5	Poodle	Maltese	Mixed
6	Unknown		Unknown
7	Golden Retriever	Poodle	Mixed

[Reeza]

Check if the second column is missing, if it is, use Column1, otherwise set it to Mixed.

if missing(column2) then desired_column = column1;
else desired_column = 'Mixed';

[advmsj]

---

@Reeza wrote:
Check if the second column is missing, if it is, use Column1, otherwise set it to Mixed.

if missing(column2) then desired_column = column1;
else desired_column = 'Mixed';

---

Thank for you this solution. What the missingness in column 2 is denoted with a 0. How does the code change?

[Reeza · Accepted solution]

Then you would check if the code was 0 rather than missing.

---

@advmsj wrote:

---

@Reeza wrote:
Check if the second column is missing, if it is, use Column1, otherwise set it to Mixed.

if missing(column2) then desired_column = column1;
else desired_column = 'Mixed';

---

Thank for you this solution. What the missingness in column 2 is denoted with a 0. How does the code change?

---