creating a percentage graph bar

[Doris Rivera]

Dear Statalist

I would like to make a graph bar but with percentages instead of the count. However, I am interested that those percentages to be 100% for each bar in total. In fact, I would like to mimic the percentage in the table below. Adding the option "percent" to the graph I´ll show below gives the percentages over the whole quantity, which is what I do not want to.
Is there any simple way to do it?

Thanks in advance.

tab x1 year, col

           |               year
        x1 |      2014       2017       2020 |     Total
-----------+---------------------------------+----------
         1 |     1,775      2,170      2,756 |     6,701 
           |     27.58      28.21      32.03 |     29.48 
-----------+---------------------------------+----------
         2 |     4,660      5,522      5,848 |    16,030 
           |     72.42      71.79      67.97 |     70.52 
-----------+---------------------------------+----------
     Total |     6,435      7,692      8,604 |    22,731 
           |    100.00     100.00     100.00 |    100.00

graph bar (count) x1, over(x2) over(year) asyvar stack blabel(bar, box bfcolor(white) pos(center) format(%4.2f) size(medium)) legend(off) ytitle("") title("from 2014")

* Example generated by -dataex-. To install: ssc install dataex
clear
input float(year x1 x2)
2014 1 1
2014 1 1
2014 2 2
2014 1 1
2014 2 2
2014 2 2
2014 1 1
2014 2 2
2014 2 2
2014 2 2
2014 . .
2014 . .
2014 . .
2014 2 2
2014 2 2
2014 2 2
2014 2 2
2014 2 2
2014 2 2
2014 2 2
2017 . .
2017 1 1
2017 2 2
2017 . .
2017 2 2
2017 2 2
2017 . .
2017 2 2
2017 1 1
2017 1 1
2017 2 2
2017 2 2
2017 2 2
2017 1 1
2017 2 2
2017 2 2
2017 2 2
2017 1 1
2017 1 1
2017 1 1
2017 2 2
2020 2 2
2020 1 1
2020 2 2
2020 1 1
2020 . .
2020 . .
2020 1 1
2020 1 1
2020 1 1
2020 2 2
2020 2 2
2020 2 2
2020 2 2
2020 2 2
2020 2 2
2020 2 2
2020 1 1
2020 2 2
2020 2 2
2020 2 2
2020 2 2
end

[Nick Cox]

In your data example x1 and x2 are always equal and your table command doesn't involve x2 while your graph command does. So, I am a little unclear what you want.

My prejudices include an idea that stacked designs are oversold. The fact that percents add to 100 is a matter of definition and it can be hard to follow categories on a stacked plot that are rare or non-existent. Legends are also an evil that should be avoided whenever possible. For all those reasons I favour what might be called a multiple bar chart instead.

See gr0061_2 for the latest files, https://www.stata-journal.com/articl...article=gr0066 for the fullest write-up and https://www.statalist.org/forums/for...updated-on-ssc for a quick overview.


SJ-20-3 gr0066_2 . . . . . . . . . . . . . . . . Software update for tabplot
(help tabplot if installed) . . . . . . . . . . . . . . . . N. J. Cox
Q3/20 SJ 20(3):757--758
added new options frame() and frameopts() allowing framing
of bars and so-called thermometer plots or charts

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


Here's some token code for your data example.

set scheme s1color 
tabplot x1 year, percent(year) separate(x1) yreverse showval bar1(fcolor(red*0.2) lcolor(red))  bar2(lcolor(blue) fcolor(blue*0.2))

The plot can be extended to three handles very easily using something like

tabplot x1 x2, by(year) 

tabplot x1 year, by(x2)

and percents can be defined any way you wish.

[Doris Rivera]

Dear Nick Cox, thanks for your help. I understand that it is confusing to have x2 in the graph but not in the table. In fact, x2 was just something I did in order to get within each year bar, the differentiation with respect to the categories of x1. This tabplot is everything I need (and more). Thanks for the suggestions.

[Nick Cox]

Pleased it helped. Naturally the percents of two exclusive categories sum to 100, and you only need one of them. Hence you could also do something like

egen toshow = mean(100 * (x1 == 1)), by(year)

graph bar (mean) toshow, over(year) ytitle(% of whatever) blabel(total, format(%2.1f))