Generate dates that correspond to the 1st Tuesday of each month

[Francois Durant]

Hi,

The Australian RBA meets the first Tuesday of each month 11 times a year (every months but January). I would like to generate these dates from now on and for the next 5 years. I have tried with the "date"command with no success. Please help!

[Nick Cox]

date() is a function, not a command. But more importantly we can't say what is wrong with your code if you don't show it.

The last day of the month was the topic of a recent riff in the Stata Journal https://journals.sagepub.com/doi/abs...urnalCode=stja Essentially, the same techniques carry over, including

Know your functions. Here dow() is key. It returns 0 to 6 for Sunday through Saturday, so we want 2s. Sometimes the first day of the month is a Tuesday, but otherwise we need to add a number between 1 and 6.

Use a sandbox dataset -- and/or Mata -- to play around with simple cases.

Get slow but sure solutions first before you try to make it smarter.

clear
set obs 11
gen firstday = mdy(_n + 1, 1, 2019)
format firstday %td

gen dow = dow(firstday)

* slow but sure; there are at worst 7 distinct cases 

gen firstTue = firstday if dow == 2
replace firstTue = firstday + 1 if dow == 1
replace firstTue = firstday + 2 if dow == 0
replace firstTue = firstday + 3 if dow == 6
replace firstTue = firstday + 4 if dow == 5
replace firstTue = firstday + 5 if dow == 4
replace firstTue = firstday + 6 if dow == 3

* did we get that right? 
assert dow(firstTue) == 2

* shorter solutions 
gen firstTue2 = cond(inrange(dow, 0, 2), firstday + 2 - dow, firstday + 9 - dow)
gen firstTue3 = firstday + mod(9 - dow, 7)

format firstTue* %td 

assert firstTue==firstTue2 & firstTue==firstTue3 

l, sep(0) 
     +-----------------------------------------------------+
     |  firstday   dow    firstTue   firstTue2   firstTue3 |
     |-----------------------------------------------------|
  1. | 01feb2019     5   05feb2019   05feb2019   05feb2019 |
  2. | 01mar2019     5   05mar2019   05mar2019   05mar2019 |
  3. | 01apr2019     1   02apr2019   02apr2019   02apr2019 |
  4. | 01may2019     3   07may2019   07may2019   07may2019 |
  5. | 01jun2019     6   04jun2019   04jun2019   04jun2019 |
  6. | 01jul2019     1   02jul2019   02jul2019   02jul2019 |
  7. | 01aug2019     4   06aug2019   06aug2019   06aug2019 |
  8. | 01sep2019     0   03sep2019   03sep2019   03sep2019 |
  9. | 01oct2019     2   01oct2019   01oct2019   01oct2019 |
 10. | 01nov2019     5   05nov2019   05nov2019   05nov2019 |
 11. | 01dec2019     0   03dec2019   03dec2019   03dec2019 |
     +-----------------------------------------------------+

[Francois Durant]

Thank you so much Nick, it does really help!