Creating a histogram with colors by a certain variable

[Hugo Denis]

Dear Statalisters,

I hope you are all fine. I'm beginning with descriptive statistics and I'd like to have a particular histogram for my project, to describe my results in a more visual way than in a table. Please find my data below :

* Example generated by -dataex-. For more info, type help dataex
clear
input float(country percent) long ind
1  .7591934 1
3 4.5177755 2
2  4.839139 3
3 2.1726158 4
4  2.386139 5
3  .4601172 6
5  31.46051 7
5  49.2879 8
1  3.964197 9
2 1.5125806 1
2  7.757139 2
3 3.7730794 3
1  .1026992 4
4 1.0024335 5
1 .25116652 6
1  74.40892 7
4 4.4953227 8
3  6.351722 9
3  2.008643 1
1  6.153969 2
5  3.598485 3
5 2.1380177 4
2  60.14992 5
3  .0933632 6
1 12.157224 7
end

As you can see, country is an identifier and each number represent a country. percent is the variable I'd like to have on my y-axis. I know histograms usually use frequencies or densities but is there a way I can plot the percentage for each industry denoted by the variable ind? In addition, I'd like to have different colors for each country denoted by the variable country. What I would like should look like something like this:



Each color should represent one country and each bar within each color should be the percentage of industry x of country y.



I hope I was clear enough. Thanks a lot for the future advices.

Best regards,

Hugo

[Nick Cox]

You have duplicates on country ind in the example dataset.

I think your one-way layout won't work nearly as well as your graph example.

Here is an alternative using tabplot from the Stata Journal. I've just summed over duplicates.

* Example generated by -dataex-. For more info, type help dataex
clear
input float(country percent) long ind
1  .7591934 1
3 4.5177755 2
2  4.839139 3
3 2.1726158 4
4  2.386139 5
3  .4601172 6
5  31.46051 7
5  49.2879 8
1  3.964197 9
2 1.5125806 1
2  7.757139 2
3 3.7730794 3
1  .1026992 4
4 1.0024335 5
1 .25116652 6
1  74.40892 7
4 4.4953227 8
3  6.351722 9
3  2.008643 1
1  6.153969 2
5  3.598485 3
5 2.1380177 4
2  60.14992 5
3  .0933632 6
1 12.157224 7
end

duplicates list country ind , sepby(country ind)

tabplot country ind [iw=percent], scheme(s1color) showval(format(%3.2f)) separate(country)

You naturally may swap axes, change the display format or change the colours. Note that red and green together is not a good idea.