_N equals the number of observations in the dataset. To see, try display _N. On the other hand, _n refers to the current observation number.

We can generate variables using _N and _n. An example: if the dataset has 5 observations and we run:

gen var1 = _N
gen var2 = _n

then, var1 will equal 5 for all observations while var2 will equal 1 for the first observation, 2 for the second, 3 for the third, and so on.

We called _n the current observation number because it delicately depends on the current sorting order of the dataset. So, if you use _n to create another variable (such as creating an ID variable), sort your dataset on a unique combination of variables before setting the ID variable equal to _n.

This is a good place to introduce indexing in Stata. We can use [] alongwith variable name to do this. For example: gen first_obs = total_covaxin[1] creates a variable first_obs that equals a constant value: the value of total_covaxin for the first observation in the dataset.

Since _n refers to the current observation number, it can be used to index by row numbers. For example, [_n-1] will refer to the observation before the current one; for the second observation, [_n-`] will equal 2 - 1 = 1, and so will refer to the first observation. So, running gen total_covaxin_prev = total_covaxin[_n-1] will create a variable that equals total_covaxin of the previous observation. Note that total_covaxin_prev is missing for the first observation, which is because Stata doesn’t have a “0” observation to refer to when _n equals 1.