Bar Graph, two variables

[Francesco Pinto]

Dear Statalists,

I am a beginner in Stata so apologies for this simple question:

I have two categorical variables (LifeSat, JobS_overall) both measuring satisfaction from 1-7 with respect to life and job.
I would like to create a bar graph showing for each category both variables together.

basically the outcome of this graph:

graph bar, over(JobS_overall)

but also with LifeSat. Sorry for the bad image but at least I clarified my goal.

Do you know how to do this?

Thank you very much!

Francesco

graph bar, over(JobS_overall)

[Justin Niakamal]

You can graph multiple yvars over a category

sysuse citytemp
graph bar (percent) heatdd cooldd, over(division)

If you post an example of your data using dataex (read the FAQ for details) you'll likely get more substantive feedback.

[Francesco Pinto]

Hi Justin,
thank you for your answer.
Below there is the data. I added pid for claritiy but it´s not relevant in my opinion.
In a panel, people were asked to report their satisfaction level with respect to life and job. Both were measured at a scale of 1 to 7.
I would like a graph like the one I posted in the picture. On the x-axis the 7 scales. On the y-axis the satisfaction rate.
It´s basically what Stata does for one variable only, and I would like to display both variable together to show the similarities in the distribution of the answer.

graph bar, over(JobS_overall)

but including LifeSat. I

* Example generated by -dataex-. To install: ssc install dataex
clear
input byte(JobS_overall LifeSat) long pid
5 . 10007857
6 . 10014608
6 . 10014608
6 . 10014608
6 5 10014608
7 6 10014608
4 6 10016813
5 . 10016848
7 . 10016848
6 6 10016872
6 . 10016872
6 6 10016872
6 6 10016872
6 6 10016872
6 6 10016872
6 6 10016872
6 6 10016872
5 6 10017933
4 5 10017933
6 . 10017968
4 . 10017968
6 . 10017968
4 5 10017992
7 5 10017992
6 7 10017992
2 4 10017992
6 5 10017992
4 . 10020209
5 . 10020209
4 . 10020233
5 . 10023526
6 4 10023526
5 5 10023526
5 5 10023526
5 5 10023526
3 5 10023526
6 5 10023526
5 5 10023526
6 5 10023526
6 6 10023526
6 . 10023569
4 . 10024646
6 . 10024646
6 5 10025804
5 5 10025804
5 6 10025804
5 . 10028005
4 . 10028005
6 . 10028005
6 . 10028005
3 5 10028005
5 5 10028005
6 5 10028005
6 5 10028005
5 5 10028005
6 . 10028005
6 5 10028005
6 4 10028005
6 5 10028005
5 5 10028005
6 5 10028005
7 5 10028005
6 . 10028382
6 . 10028382
6 . 10028382
6 . 10028382
5 . 10029133
4 . 10029133
6 6 10029133
5 5 10029133
5 6 10029133
5 . 10029133
7 . 10040404
6 7 10040404
7 . 10048219
6 . 10048219
7 7 10048219
7 7 10048219
7 7 10048219
7 6 10048219
6 6 10048219
6 6 10048219
6 6 10048219
7 . 10048243
6 . 10048243
6 . 10048243
6 . 10048243
7 6 10048243
7 6 10048243
6 5 10048243
7 3 10048243
5 . 10048243
6 5 10048243
6 5 10048243
5 5 10048243
5 6 10048243
7 . 10048278
6 . 10048278
5 . 10048278
6 6 10048278
end
label values JobS_overall rjbsat
label def rjbsat 4 "Not satis/dissatis", modify
label def rjbsat 7 "Completely satis", modify
label values LifeSat rlfsato
label def rlfsato 4 "Not satis/dissat", modify
label def rlfsato 7 "Completely satisfied", modify

[Francesco Pinto]

basically I would like to put both graphs together.

graph bar, over(JobS_overall) bar(1, fcolor(red))

graph bar, over(LifeSat) bar(1, fcolor(yellow))

[Justin Niakamal]

Not sure how you want to handle missing data. Hopefully this is enough to get you started,

tempfile temp 

preserve 
contract JobS_overall
egen sh_jobs = pc(_freq)
drop _freq

ren JobS_overall category 
save `temp'
restore 


contract LifeSat
egen sh_life = pc(_freq)
drop _freq

ren LifeSat category
merge 1:1 category using `temp'

label var sh_life "Life Satisfaction"
label var sh_jobs "Jobs Overall"

graph  bar sh_* , over(category)

[Nick Cox]

A lot of work to do here in terms of variable labels, consistent and complete value labels, and so on. But this may help:

* Example generated by -dataex-. To install: ssc install dataex
clear
input byte(JobS_overall LifeSat) long pid
5 . 10007857
6 . 10014608
6 . 10014608
6 . 10014608
6 5 10014608
7 6 10014608
4 6 10016813
5 . 10016848
7 . 10016848
6 6 10016872
6 . 10016872
6 6 10016872
6 6 10016872
6 6 10016872
6 6 10016872
6 6 10016872
6 6 10016872
5 6 10017933
4 5 10017933
6 . 10017968
4 . 10017968
6 . 10017968
4 5 10017992
7 5 10017992
6 7 10017992
2 4 10017992
6 5 10017992
4 . 10020209
5 . 10020209
4 . 10020233
5 . 10023526
6 4 10023526
5 5 10023526
5 5 10023526
5 5 10023526
3 5 10023526
6 5 10023526
5 5 10023526
6 5 10023526
6 6 10023526
6 . 10023569
4 . 10024646
6 . 10024646
6 5 10025804
5 5 10025804
5 6 10025804
5 . 10028005
4 . 10028005
6 . 10028005
6 . 10028005
3 5 10028005
5 5 10028005
6 5 10028005
6 5 10028005
5 5 10028005
6 . 10028005
6 5 10028005
6 4 10028005
6 5 10028005
5 5 10028005
6 5 10028005
7 5 10028005
6 . 10028382
6 . 10028382
6 . 10028382
6 . 10028382
5 . 10029133
4 . 10029133
6 6 10029133
5 5 10029133
5 6 10029133
5 . 10029133
7 . 10040404
6 7 10040404
7 . 10048219
6 . 10048219
7 7 10048219
7 7 10048219
7 7 10048219
7 6 10048219
6 6 10048219
6 6 10048219
6 6 10048219
7 . 10048243
6 . 10048243
6 . 10048243
6 . 10048243
7 6 10048243
7 6 10048243
6 5 10048243
7 3 10048243
5 . 10048243
6 5 10048243
6 5 10048243
5 5 10048243
5 6 10048243
7 . 10048278
6 . 10048278
5 . 10048278
6 6 10048278
end
label values JobS_overall rjbsat
label def rjbsat 4 "Not satis/dissatis", modify
label def rjbsat 7 "Completely satis", modify
label values LifeSat rlfsato
label def rlfsato 4 "Not satis/dissatis", modify
label def rlfsato 7 "Completely satis", modify

su JobS_overall
bysort JobS_overall : gen y1 =  100 * _N/r(N)
label var y1 "%"

su LifeSat 
bysort LifeSat : gen y2 = 100 * _N/r(N)
label var y2 "%"

twoway bar y1 JobS_overall, barw(0.8) bfcolor(red) name(G1, replace) xla(2/7, valuelabel) yla(0(20)60)

twoway bar y2 LifeSat, barw(0.8) bfcolor(yellow) name(G2, replace) xla(2/7, valuelabel) yla(0(20)60)

graph combine G1 G2, col(1) xcommon

[Francesco Pinto]

Thank you very much Justin. This is what I wanted to achieve although the numbers are different.

[Justin Niakamal]

Thank you very much Justin. This is what I wanted to achieve although the numbers are different.

I think that's because of the missings. Maybe this is closer.

tempfile temp 

preserve 
drop if missing(JobS_overall)
contract JobS_overall
egen sh_jobs = pc(_freq)
drop _freq

ren JobS_overall category 
save `temp'
restore 

drop if missing(LifeSat)

contract LifeSat
egen sh_life = pc(_freq)
drop _freq

ren LifeSat category
merge 1:1 category using `temp'

label var sh_life "Life Satisfaction"
label var sh_jobs "Jobs Overall"

graph  bar sh_* , over(category)

[Francesco Pinto]

Thank you very much Justin, it worked perfectly!
Thank you too Nick for your help!