Scatter Plot, Twoway plot

[John Kab]

I am making a scatter plot of changes in GDP vs changes in the ratio of old-age dependency panel data. How can I do it?
Thank you!

[Rich Goldstein]

help scatter

if this is not sufficient, please read the FAQ on how to ask questions that get useful answers

[John Kab]

It's not sufficient. To be clear, I want a plot of changes in GDP vs changes in the ratio of old-age dependency. Here's my code

twoway (scatter change_gdp change_dependency if sample_oecd!=1 , msize(*2) msymbol(circle) mfcolor(ltblue) mlcolor(black)) ///
(scatter change_gdp change_dependency if sample_oecd!=1 & (change_gdp>2|change_dependency>.2), mlabel(countrycode) msize(*2) msymbol(circle) mfcolor(ltblue) mlcolor(black)) ///
(scatter change_gdp change_dependency if sample_oecd==1 & cname_ifr!="Japan" , mlabel(countrycode) msize(*2) msymbol(diamond) mfcolor(dkorange) mlcolor(black)) ///
(scatter change_gdp change_dependency if cname_ifr=="Japan", mlabel(countrycode) msize(*2) msymbol(triangle) mfcolor(cranberry) mlcolor(black)) ///
(lfit change_gdp change_dependency, lcolor(navy)), ///
ylabel(#6, glwidth(thin) glcolor(gs12)) ///
legend(on cols(2) order(1 "All countries except OECD" 3 "OECD countries" )) ytitle("Change in GDP per capita from 1990 to 2015") xtitle("Change in the old age dependecy ratio") ///
graphregion(fcolor(white) ifcolor(white) lcolor(white) ilcolor(white) ) plotregion(fcolor(white) ifcolor(white) lcolor(white) lwidth(none) ilcolor(white) )

[Nick Cox]

I can't yet see an answerable question here. You don't give example data. On the face of it the code gives plots of change_gdp versus change_dependency. but there is no graph example either. So what's wrong?

[John Kab]

Sorry! Please find attached the graph that I want to replicate and the data.

twoway (scatter change_gdp change_dependency if sample_oecd!=1 , msize(*2) msymbol(circle) mfcolor(ltblue) mlcolor(black)) ///
(scatter change_gdp change_dependency if sample_oecd!=1 & (change_gdp>2|change_dependency>.2), mlabel(countrycode) msize(*2) msymbol(circle) mfcolor(ltblue) mlcolor(black)) ///
(scatter change_gdp change_dependency if sample_oecd==1 & cname_ifr!="Japan" , mlabel(countrycode) msize(*2) msymbol(diamond) mfcolor(dkorange) mlcolor(black)) ///
(scatter change_gdp change_dependency if cname_ifr=="Japan", mlabel(countrycode) msize(*2) msymbol(triangle) mfcolor(cranberry) mlcolor(black)) ///
(lfit change_gdp change_dependency, lcolor(navy)), ///
ylabel(#6, glwidth(thin) glcolor(gs12)) ///
legend(on cols(2) order(1 "All countries except OECD" 3 "OECD countries" )) ytitle("Change in GDP per capita from 1990 to 2015") xtitle("Change in the old age dependecy ratio") ///
graphregion(fcolor(white) ifcolor(white) lcolor(white) ilcolor(white) ) plotregion(fcolor(white) ifcolor(white) lcolor(white) lwidth(none) ilcolor(white) )

[Marcos Almeida]

I gather something like - scatter var1 var2, mlabel(var3) - may do the trick.

[John Kab]

It doesn't work

[Marcos Almeida]

Please take a look at the FAQ. There you’ll find how to share data/command/output. This is the best approach to reap a helpful reply. Saying ‘it doesn’t work’, as also underscored in the FAQ, is not much informative.

[John Kab]

Thank you! I have already solved the problem.