Trouble with formatting dates

[Hanna Thompson-Schumann]

Hello all,

I am having trouble remembering how I formatted dates to look like this:


date2 under50 over100
201306 2 2.19
201307 2.135 2
201308 2 2.19
201309 1.9 2.2
201310 1.43 2.42
201311 1.59 2.13
201312 1.615 2.39
201401 2 2.54
201402 2.3249998 2.57
201403 2.45 2.79
201404 1.38 2.695
201405 2.12 2.45
201406 2.23 2.845
201407 2 2.45
201408 1.785 2.81
201409 1.69 2.59
201410 1.9 2.53
201411 2.395 2.97
201412 2 2.54
201501 2 2.65
201502 1.81 2.96
201503 2 2.94
201504 2 2.675
201505 2 2.695
201506 2.025 2.75
201507 2 2.98
201508 1.75 3
201509 1.63 2.71
201510 2 2.71
201511 2 2.61
201512 1.42 2.52
201601 1.55 2.54
201602 1.345 2.55
201603 1.46 2.62
201604 1.4 2.485
201605 1.86 2.54
201606 1.465 2.8
201607 2 2.5
201608 2.19 2.5
201609 1.81 2.49
201610 2 2.33
201611 1.66 2.435
201612 1.825 2.62
201701 2 2.2849998
201702 1.465 2.71
201703 1.99 2.7
201704 2.54 2.62
201705 2 2.81
201706 2.37 2.71
201707 2.42 2.705
201708 2 2.905
201709 2 2.71
201710 2.12 2.77
201711 2.145 2.83
201712 2.52 2.98
201801 2.66 2.785
201802 2.81 2.76
201803 2.31 2.88
201804 2 3
201805 2.21 2.82
201806 2.565 2.81
201807 2 2.81
201808 2.345 2.98
201809 2 3
201810 2 3
201811 1.92 2.67
201812 2 2.92
201901 1.69 2.98
201902 2 2.78
201903 2.19 2.98
201904 1.685 2.89
201905 2 2.79
201906 1.7 2.79
201907 1.76 2.79
201908 1.89 2.69
201909 2 2.98
201910 1.53 2.61
201911 1.555 2.89
201912 2 2.775
202001 1.535 2.98
202002 2 2.81
202003 1.735 2.25
202004 1.31 2
202005 1 2
202006 1.0799999 2
202007 1.51 2
202008 1.29 2.2150002
202009 1.19 2
202010 1.29 2
202011 1.16 2.3
202012 1.33 2.28
202101 1.43 2.675
202102 1.29 2.71
202103 1.19 2.54
202104 1.57 2.975
202105 2 2.98
202106 2 3
202107 2 3
end

twoway line median median2 date2, title("One Year Ahead Median Expected Earnings Growth", size(medium)) lcolor(blue%50 red%50) ytitle("") xtitle("") ylabel(0 .5 1 1.5 2 2.5 3 3.5 4 4.5 5,grid) xlabel(,format(%tdCY)) legend(size(small) pos(6) row(1) lab(1 "Income under 50k") lab(2 "Income over 100k")) graphregion(fcolor(white))

this is the code but it is not working. any help would be appreciated.

[Nick Cox]

Please use .png not .pdf attachments for images. This is explained at FAQ Advice #12.

The brute fact is that monthly dates like this are good for people, who can read them quite easily, but lousy for analysis, as each series of 12 monthly dates taken literally (which means numerically) implies 11 gaps of 1 and 1 of 89, which can only look bizarre on a graph. So integers like 201712 and 201801 just aren't going to be decoded inside Stata as you want.

Also, the function monthly() is not smart enough to deal with them, even when you present it with a string equivalent.

. di monthly("202211", "YM")
.

Some surgery is needed as in

clear
input date2 under50 over100
201307 2.135 2
201308 2 2.19
201309 1.9 2.2
201310 1.43 2.42
201311 1.59 2.13
201312 1.615 2.39
201401 2 2.54
201402 2.3249998 2.57
201403 2.45 2.79
201404 1.38 2.695
201405 2.12 2.45
201406 2.23 2.845
201407 2 2.45
201408 1.785 2.81
201409 1.69 2.59
201410 1.9 2.53
201411 2.395 2.97
201412 2 2.54
end

gen mdate = ym(floor(date2/100), mod(date2, 100))
format mdate %tm


list mdate, sep(6)

     +---------+
     |   mdate |
     |---------|
  1. |  2013m7 |
  2. |  2013m8 |
  3. |  2013m9 |
  4. | 2013m10 |
  5. | 2013m11 |
  6. | 2013m12 |
     |---------|
  7. |  2014m1 |
  8. |  2014m2 |
  9. |  2014m3 |
 10. |  2014m4 |
 11. |  2014m5 |
 12. |  2014m6 |
     |---------|
 13. |  2014m7 |
 14. |  2014m8 |
 15. |  2014m9 |
 16. | 2014m10 |
 17. | 2014m11 |
 18. | 2014m12 |
     +---------+

As always the essential reading is

help datetime
help datetime display formats

But all that said, your graph shows monthly data for about 8 years and on that scale years are definitely intervals, not points. So what is the reader supposed to think about the tick for 2013? Does it represent the start, the end, or the middle of 2013? A better approach is explained at https://www.stata-journal.com/articl...article=gr0030