pweight on twoway bar graphs

[Denat Ephrem Negatu]

I'm working with the World Bank Enterprise Survey dataset. I'm trying to plot a twoway bar graph with confidence intervals. How can I add pweight to the graph? Would converting pweight to fweight make sense in this case? Or is there another way to incorporate pweights into a twoway bar graph?

[Andrew Musau]

collapse allows sampling weights, so collapse the data specifying the weights and graph. Here is an example using gr bar.

sysuse auto, clear
set scheme s1mono
gr bar mpg [pw=price], over(rep78) blab(total, format("%3.2f")) ///
ytitle("`:var lab mpg'") ysc(r(0 25.5)) saving(gr1, replace) title("Raw data")

*USING COLLAPSE
collapse mpg [pw=price], by(rep78)
gr bar mpg, over(rep78) blab(total, format("%3.2f")) ///
ytitle("`:var lab mpg'") ysc(r(0 25.5)) saving(gr2, replace) title("Using collapse")

*COMPARISON
gr combine gr1.gph gr2.gph

Res.:

[Denat Ephrem Negatu]

Thank you Andrew. However, I am collapsing so as to also get the standard errors inorder to incorporate a confidence interval on the graphs. My collapse command is as follows

collapse (mean) mean= var_1 (sd) sd=var_1 (count) n=var_1 , by (catagory year)

in the above command, or along anything with standard deviation, pweights don't seem to be allowed. Is there any way to address this?

[Denat Ephrem Negatu]

Thank you Andrew. However, I am collapsing so as to also get the standard errors inorder to incorporate a confidence interval on the graphs. My collapse command is as follows

collapse (mean) mean= var_1 (sd) sd=var_1 (count) n=var_1 , by (catagory year)

in the above command, or along anything with standard deviation, pweights don't seem to be allowed. Is there any way to address this?

[Andrew Musau]

You can do several collapses, but I do not fully understand the problem. It may be possible to compute the standard errors using -svy- estimation, but I am not too knowledgeable in this area to advise appropriately. I would suggest starting a new thread explaining in full details what you want to do and attaching a data example so that those who have expertise in the area can advise you.