Seek help for concentration curve

[Kyaw Show]

Dear Sir/Madam, colleagues,

I tried to plot a concentration curve with the following commend from some literatures but did not get a figure.
Please kindly correct and guide me further.

glcurve outcome1 [aw=wgt], glvar(g1) pvar(p1) sortvar(v191) lorenz nograph
glcurve outcome2 [aw=wgt], glvar(g2) pvar(p2) sortvar(v191) lorenz nograph
glcurve outcome3 [aw=wgt], glvar(g3) pvar(p3) sortvar(v191) lorenz nograph
glcurve outcome4 [aw=wgt], glvar(g4) pvar(p4) sortvar(v191) lorenz nograph

label variable g1 "g1"
label variable g2 "g2"
label variable g3 "g3"
label variable g4 "g4"

graph twoway (line g1 p1, sort yaxis(1 2)) ///
(line g2 p2, sort yaxis(1 2)) ///
(line g3 p3, sort yaxis(1 2)) ///
(line g4 p4, sort yaxis(1 2)) ///
(function y=x, range(0 1) yaxis(1 2)) ///
, aspect(1) xtitle("cumulative proportion share, p") ///
title("Lorenz curve comparison") ///
ytitle("cumulative proportion wealth") ///
legend(label(1 "outcome1") label(2 "outcom2") label (3 "outcome3") label(4 "outcome4") ///
saving(test.gph)

Where outcome 1-4 are the dichotomous (0 1) and v191 is the continuous variable.
I tried to include all 4 outcomes in single figure as the inquality variable is the same.
Thanking in advance.

SHOW

[Dung Le]

Hi Show,

It would be helpful if you can provide an example of your data. Saying "did not get a figure" is somewhat vague. Another way to plot a CC is to use -lorenz- command written by Ben Jann (type ssc install lorenz for installation).

Hope this helps.

[Kyaw Show]

Dear Dung,

Thank you for your reply. I am doing an inequality analysis using DHS data.
Outcome 1 would be like healthcare seeking and v191 is the wealth index in continuous variable.
I hope this is what you mean. Thanks. Will also explore Ben Jann commend too.

SHOW

[Dung Le]

Actually, what I meant is an example of data and normally people here use -dataex- command to do so. See "12. What should I say about the commands and data I use" in FAQ for more detail. However, If y is your outcome then try this.
Note: title, labels, xtitle and ytitle depend on your purpose so I left those options blank.

lorenz y, pvar(v191) ///
    graph(aspectratio(1) xlabels(, grid) overlay legend(cols(1)) noci ///
    title("") labels("") lpattern(dash) xtitle("") ytitle(""))

[Kyaw Show]

Dear Dung,

Thank you very much for your kind guidance.
It worked with the lorenz commend.

[Kyaw Show]

Dear Dung and all,

Please kindly also guide me how to get 95%ci for concentration index.
I used the following commend to get concentration index.
Thanks

conindex outcome [pw=wgt], rankvar(variable) bounded limits(0 1) erreygers cluster(v001)

SHow

[Dung Le]

Hi Show,

Computing SE is quite straightforward, try this

conindex outcome [pw=wgt], rankvar(variable) limits(0 1) bounded erreygers  cluster(v001)  
* Store standard error of CI
sca SE = r(CIse)
* Generate lower bound of 95% CI
gen lb =  CI - invttail(e(df_r),0.025)*SE
* Generate a upper bound of CI
gen ub = CI + invttail(e(df_r),0.025)*SE

[Kyaw Show]

Dear Dung,

Thank you very much.

Show