Generating New Observations

[MuhammetAli Cansi]

Hello everyone,

I want to learn if it is possible to generate new observations in the same variables with using the existing variables.
Assume the following example: Time series data from 1980-2022 and the variable for the years is "Year" and there is "Month" variable 1-12 for each year. Now further assume that we have variables var_1........var_20 .
I want to create 2023 under the year with 12 observation 1-12 under month. and for all the other variables var_1....var_20 I want to generate the values as the sum of values from 2009 and 2010. I thought it should be something like that.

gen obs(2023) = if Year==2009 + if Year=2010

Of course such code does not exist and does not work I gave it as an indication to make my question more understandable. Is it possible to do that. I can do it in new variable but I have no idea how can I generate new observations from the existing observations. This example is just for the general case. I have not seen any related topic before so if there are please inform me so that I will follow that topic.

[Nick Cox]

If I understand this correctly you have 43 * 12 = 516 observations. You want an extra 12,

set obs 528
replace Year = 2023 in -12/L
replace Month = _n - 516 in -12/L

Otherwise that sounds a very strange thing to want, but this seems to be what you are asking

forval j = 1/20 {
      su var_`j' if inlist(Year, 2009, 2010), meanonly
      di _n  "var_`j'    "  r(sum)
      replace var_`j' = r(sum) if Year == 2023
}

If you want the values for e.g. January 2023 to be the sum of January 2009 and January 2010, then that could be done, but i will not write code at this point.

[MuhammetAli Cansi]

If I understand this correctly you have 43 * 12 = 516 observations. You want an extra 12,

set obs 528
replace Year = 2023 in -12/L
replace Month = _n - 516 in -12/L

Otherwise that sounds a very strange thing to want, but this seems to be what you are asking

forval j = 1/20 {
su var_`j' if inlist(Year, 2009, 2010), meanonly
di _n "var_`j' " r(sum)
replace var_`j' = r(sum) if Year == 2023
}

If you want the values for e.g. January 2023 to be the sum of January 2009 and January 2010, then that could be done, but i will not write code at this point.

Thank you so much that is exactly what I ask. I understand it seems strange however for some of the data in the long format that I am dealing with I need to create generated observations using the existing observations. That's why I need a way to be able to generate them. I got the idea and will be able to do rest of it my self you have provided enough detail. Thanks again.