What if we want to know the number of vaccinations done on April 2 or the average number of daily covishield vaccinations done in March in Jalandhar, Punjab?

If qualifier

We can use the if qualifier with our commands to get the relevant numbers for particular group of observations. Important: I called it a “qualifier” and not a “command” because it helps subsets/qualifies certain observations for further use in a Stata command. Lets compute the average, district-wise number of daily covishield vaccinations done in March throughout India:

summarize total_covishield if date_month == 3, detail

It looks like the average number of daily covishield vaccinations in a district was 35486 in March. The distribution is quite spread out. The 5th percentile is 2161 whereas teh 95th percentile is over 1,00,000. Of course, these figures haven’t been normalized for district population, i.e., looking at absolute number of vaccinations does not take into account that vaccinations could be lower for districts with lower populations.

Anyway, lets get back to what we ran. The additional piece is if date_month == 3. We are telling Stata to only consider data from the month of March. Also note that the , option1 option2 ... segment of the code comes after the ‘if statement’ and not before.

And/Or

What if you want to now look at the average daily vaccinations in Punjab’s districts in March? That is, instead of all Indian districts, we want to restrict the sample to Punjab districts. So, we now have two statements we want to subset on: date_month == 3 and lgd_state_name == "punjab". Sorry my markdown (tool in which this webpage is written) skills are still not great enough, so you will need to imagine a venn diagram with two sets: set of all observations with date_month == 3 and another set of observations with lgd_state_name == "punjab". Do we want an intersection or a union of these sets?

An intersection! We want to look at the data pertaining to Punjab districts in March. An intersection means “and”, i.e., both the conditions should be satisfied. Stata does not understand the english word “and”, instead we need to use “&”. So, how will our if statement look? if date_month == 3 & lgd_state_name == "punjab". So, we run:

summarize total_covishield if date_month == 3 & lgd_state_name == "punjab", detail

For completeness, let’s come back to the question posed at the top. Try figuring out how you would do it.

What I’d try: First tabulate Punjab districts to check the capitalization and spelling of Jalandhar and then run the actual command.

tab lgd_district_name if lgd_state_name == "punjab"
summarize total_covishield if date_month == 3 & lgd_state_name == "punjab" & district == "jalandhar", detail

What if you want to look at average daily covishield vaccinations in Punjab and Haryana districts (combined) in March?

We answer this in two steps:

1. Lets forget we care about March and focus on all the months. After this, we have two sets of observations we care about: Punjab observations and Haryana observations. Intersection or union of these sets? Union! That is, observations belonging to either Punjab or Haryana. Similar to “and”, Stata does not understand the english word “or”, instead we need to use “|”. So, our if statement would like: if lgd_state_name == "punjab" | lgd_state_name == "haryana"

2. We want to subset for Punjab and Haryana observations in March. We cannot write if date_month == 3 & lgd_state_name == "punjab" | lgd_state_name == "haryana" because it is unclear if this statement means “March and Punjab, or Haryana” or “March, and Punjab or Haryana”. Stata will execute the former (i.e., Punjab obs only from March while all Haryana observations) whereas we desire the latter. Similar to the BODMAS rule for basic math functions, this calls for the use of parentheses to make sure Stata does exactly what you want it to do. The correct way to write, therefore, is if date_month == 3 & (lgd_state_name == "punjab" | lgd_state_name == "haryana"). Another correct way to write is (if date_month == 3 & lgd_state_name == "punjab") | (date_month == 3 & lgd_state_name == "haryana"). Make sure you understand why these two are equivalent. For completeness, we would run:

summarize total_covishield if date_month == 3 & (lgd_state_name == "punjab" | lgd_state_name == "haryana")

Always good to sketch out a venn diagram to ensure you understand what conditional statement you need to use!