histogram over age groups?

[River Huang]

Dear All, I find this question here. The data set is:

* Example generated by -dataex-. To install: ssc install dataex
clear
input double(age cognlab)
48 3
50 2
59 2
62 2
50 2
48 2
63 2
55 1
69 2
70 3
83 2
81 1
83 1
58 1
61 2
47 3
47 3
59 1
67 1
59 2
53 2
56 2
51 2
49 2
57 1
64 2
52 1
47 1
55 2
59 2
59 2
68 2
56 1
53 1
59 1
55 2
48 1
47 3
46 1
43 2
47 1
49 1
66 2
46 2
45 2
63 1
62 1
47 3
57 1
64 2
58 1
72 1
56 2
61 1
45 2
41 3
54 2
56 2
67 1
59 1
59 2
48 1
52 2
49 2
56 2
73 2
70 1
61 2
80 1
76 1
63 1
50 2
58 2
50 3
52 2
66 1
59 2
60 1
54 2
55 2
48 2
75 2
66 2
59 2
58 2
74 1
62 2
60 2
64 2
67 1
51 2
51 2
53 2
70 1
70 2
61 2
61 2
67 1
61 1
56 1
end

1. The end is to draw a graph like below:
   
2. Age in on the horizontal axis, with three groups. The first group is 45-60, the second group is 60-75, and the third group is age > 75.
3. The variable on the vertical axis is `cognlab', which has only three values, i.e., 1,2, and 3.
4. Any suggestions are appreciated.

[Marcos Almeida]

I could not read the link, for it is not in English. That said, it seems you need to deal with stacked bar graphs.

[River Huang]

Dear Marcos, The original question is in Chinese. I guess so too (stacked bar) but the problem is how?

[Nick Cox]

This would often be called a stacked (or subdivided) bar chart. You could do it with graph bar, although for this particular application I am more familiar with catplot (SSC). I would suggest that tabplot (Stata Journal) is also competitive here.

gen age2 = 15 * floor(age/15)
foreach age in 30 45 60 75 {
    local AGE = `age' + 14
    label define age2 `age' "`age'-`AGE'", modify
}
label val age2 age2

catplot cognlab age2, percent(age2) stack recast(bar) asyvars yla(, ang(h)) ///
legend(order(3 2 1) pos(3) col(1)) ///
bar(3, bcolor(blue*0.6)) ///
bar(2, fcolor(blue*0.2) lcolor(blue*0.6)) ///
bar(1, fcolor(red*0.2) lcolor(red*0.6)) b1title(Age (years)) name(G1, replace)

tabplot cognlab age2, percent(age2) yla(, ang(h)) ///
showval ///
xtitle(Age (years)) name(G2, replace) ///
subtitle(% in age group) ///
yreverse ///
separate(cognlab) ///
bar3(bcolor(blue*0.6)) ///
bar2(bfcolor(blue*0.2) blcolor(blue*0.6)) ///
bar1(bfcolor(red*0.2) blcolor(red*0.6))

[Marcos Almeida]

River Huang My comment in #2 was basically to confirm that you wanted the stacked bar graphs, as I guessed, but not some histograms, as stated in the title.

Being this so, in spite of Nick having already given a couple of great strategies, I just wish to share this text since it shows how to produce such graphs with Stata's ordinary commands.

Hopefully it helps.

[Nick Cox]

The age binning is avoidable. This works intelligibly with the sample given, but not so well with much larger samples.

* ssc inst stripplot 
stripplot age, over(cognlab) stack height(0.4) ms(Sh)

[River Huang]

Dear Marcos, Thanks for the suggestion. I will pass this information to the questioner.

[River Huang]

Dear Nick, Thanks for the suggestion. I will pass this information to the questioner.

[River Huang]

Dear Nick, I tried your code and found the following message

. tabplot cognlab age2, percent(age2) yla(, ang(h)) ///
> showval ///
> xtitle(Age (years)) name(G2, replace) ///
> subtitle(% in age group) ///
> yreverse ///
> separate(cognlab) ///
> bar3(bcolor(blue*0.6)) ///
> bar2(bfcolor(blue*0.2) blcolor(blue*0.6)) ///
> bar1(bfcolor(red*0.2) blcolor(red*0.6))
option yreverse not allowed
r(198);

Any idea?

[Nick Cox]

So, you are not using the latest version. Pay attention, 007. https://www.imdb.com/title/tt0879840/

search tabplot

SJ-17-3 gr0066_1 . . . . . . . . . . . . . . . . Software update for tabplot
(help tabplot if installed) . . . . . . . . . . . . . . . . N. J. Cox
Q3/17 SJ 17(3):779
added options for reversing axis scales; improved handling of
axis labels containing quotation marks

SJ-16-2 gr0066 . . . . . . Speaking Stata: Multiple bar charts in table form
(help tabplot if installed) . . . . . . . . . . . . . . . . N. J. Cox
Q2/16 SJ 16(2):491--510
provides multiple bar charts in table form representing
contingency tables for one, two, or three categorical variables

SJ-4-2 gr0004 . Speaking Stata: Graphing categorical and compositional data
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . N. J. Cox
Q2/04 SJ 4(2):190--215 (no commands)
discusses graphical possibilities for categorical and
compositional data

[River Huang]

Dear Nick, I have updated the -tabplot- package as follows.

However, same error message appeared.

. gen age2 = 15 * floor(age/15)

. foreach age in 30 45 60 75 {
  2.     local AGE = `age' + 14
  3.     label define age2 `age' "`age'-`AGE'", modify
  4. }

. label val age2 age2

. 
. tabplot cognlab age2, percent(age2) yla(, ang(h)) ///
> showval ///
> xtitle(Age (years)) name(G2, replace) ///
> subtitle(% in age group) ///
> yreverse ///
> separate(cognlab) ///
> bar3(bcolor(blue*0.6)) ///
> bar2(bfcolor(blue*0.2) blcolor(blue*0.6)) ///
> bar1(bfcolor(red*0.2) blcolor(red*0.6))
option yreverse not allowed
r(198);

[Nick Cox]

Presumably, the old program code is still in memory.

discard

or

program drop tabplot

[River Huang]

Dear Nick, Unfortunately, it doesn't work, either. (same error message!) I uninstall tabplot and re-install tabplot, but with no luck.

[Nick Cox]

I can't easily help further. The code worked for me -- otherwise #4 was impossible -- and the version on my machine is what you can see by

type http://www.stata-journal.com/software/sj17-3/gr0066_1/tabplot.ado

You need to check

which tabplot, all

as somehow you have left an older version for Stata to use.

[Nick Cox]

Also, as said, update from the Stata Journal , not from SSC.