Generating unique ID's for my observations

[rachita vig]

Hi everyone,

I have data of the following type for which I want to generate unique ID's:

School_ID Number of classrooms in grade 9 Number of classrooms in grade 8
1101 2 1
1102 1 0

I want to generate a variable by the name of Class_ID which should contain an unique ID for each classroom of each grade.

For example: For School with School_ID 1101 -- I want the class ID for the first classroom of grade 9 to be like 11019A (where in the first four digits are the School_ID, next digit is what grade whether 9 or 8, and last digit is the unique identifier for that classroom denoted by letter A).

Similarly, for the another classroom in grade 9, the ID should be 11019B.

Is there a code which can help me to do so?

Kindly please suggest!

Thank you very much,

Rachita

[Matt Warkentin]

Hi Rachita,

Based on the two lines of example code, it looks like you only have one line of data per school and multiple variables within a school (i.e. wide data). In order to achieve what you are asking you need to convert your data from wide to long (see Stata command -reshape-). This will allow multiple classrooms within a school with the same school ID, and each classroom will be a unique observation. Is this what you wanted? An example of this would look something like:

schoolID grade classID
1101 9 11019A
1101 9 11019B
1101 8 11018A
1102 9 11029A

[rachita vig]

Hi Matt,

Thanks for replying! So what have here is in wide format :-

School_id grade8 grade9
1101 1 2
1102 0 1

I used -reshape- as you suggested by: reshape long grade, i(School_id) j(class_id)

So I got this:

School_id class_id grade
1101 8 1
1101 9 2
1102 8 0
1102 9 1

But, this is not what I want. How did you get yours?

Kindly help.

[Nick Cox]

Rachita: Please read and act on FAQ Advice #12 on showing data examples and code. I see no classroom variable in your example, but this is I think similar in spirit to Matt's solution.

clear 
input School_id grade8 grade9 str1 classroom 
1101 1 2  A 
1102 0 1  A 
end 

reshape long grade, i(School_id) j(Class_id)

egen ID = concat(School_id Class_id classroom) 

list 


     +-------------------------------------------------+
     | School~d   Class_id   grade   classr~m       ID |
     |-------------------------------------------------|
  1. |     1101          8       1          A   11018A |
  2. |     1101          9       2          A   11019A |
  3. |     1102          8       0          A   11028A |
  4. |     1102          9       1          A   11029A |
     +-------------------------------------------------+

[Andrew Musau]

If grade defines a count of the number of classrooms, you should delete observations for which grade=0 in #3. In the absence of a classroom variable in Nick's example, #2 from the following link may be useful

https://www.statalist.org/forums/for...ns-in-alphabet