0

My dataset looks like this:

Country year poverty rate sales
Austria 1950 0.54 142
Austria 1951 0.32 12441
Austria 1952 0.32 12441
Bangladesh 1950 0.11 142123123
Bangladesh 1951 0.52 1234
Bangladesh 1952 0.32 12441
Sri Lanka 1950 0.95 4215
Sri Lanka 1951 0.21 142421
Sri Lanka 1952 0.32 12441

I want to do tsset so that I can (for example) create a new variable for change in sales per year for each country. When I try to do tsset country year, I see "repeated time values within panel". How can I create a new variable that is change in sales per year for each country and year? I have more variables so I would want to be able to specify the variable.

Nick Cox
  • 35,529
  • 6
  • 31
  • 47
aszdaasdasd
  • 3
  • 2

1 Answers1

0

country looks like a string variable from here, but if it were then

 tsset country year

would fail for that reason. So, suppose country is a numeric variable with value labels. Then it is essential to follow up the report of repeated observations with say

 duplicates list country year 

 duplicates tag country year, gen(tag) 

 edit if tag

Then the next step depends on what you see, for example:

  1. The duplicates are just junk with missing values on one of those variables. drop the junk.

  2. Accidental duplicate observations. drop the duplicates.

  3. Something more serious.

See also FAQ https://www.stata.com/support/faqs/data-management/repeated-time-values/

Nick Cox
  • 35,529
  • 6
  • 31
  • 47