Overlapping labels of pie chart

[Simay Kizilkaya]

Dear Statalisters,

I have to report several pie charts in a study. However, I face with the problem of overlapping labels. How can I add line connectors to the pie chart which are outside the pie chart but connects the shares? I am attaching an example of the problem. Thank you in advance.


Sincerely,

[Nick Cox]

Please note our requests to give a data example, to show the code you used, and to present graph attachments as .png not .gph -- all explicit at https://www.statalist.org/forums/help#stata.

I read off the following data. No doubt the text labels could be made more informative. For example, "coal imp" is perhaps coal imports and "asph" may mean "asphalt".

* Example generated by -dataex-. To install: ssc install dataex
clear
input str11 source float percent
"renewable" 41.91
"natural gas" 18.9
"lignite" 16.09
"coal imp" 20.8
"fuel oil" .3191
"asph coal" .8003
"coal" 1.178
end

It's hard to think of any advantage to pie charts beyond their being widely familiar. The message that components add to 100% is no doubt educational or reassuring for young children, but not otherwise.

Any problem with placement of labels is solved completely by using a bar chart -- which also does away with the awkward legend. Several small choices remain. For example, for graphics that have table flavour, I often like to put the magnitude axis at the top, but that is personal taste (more on that at https://www.stata-journal.com/articl...article=gr0053).

The code here is just one of many similar choices. I would for example suggest standardising on 1 or 2 decimal places.

graph hbar (asis) percent, over(source, sort(1) descending) scheme(s1color) blabel(bar) bar(1, lcolor(red) fcolor(red*0.2)) ysc(r(0 44) alt) ytitle(percent of total)

[Andrés Lahur Talavera Cuya]

How to make a graph similar to this:

This is what I have progressed:

* Example generated by -dataex-. For more info, type help dataex
clear
input byte d_ipcm float ipcm
 1 142.46667
 2  252.4869
 3  344.0508
 4  432.8683
 5  531.2717
 6  651.0728
 7  801.3522
 8 1011.8083
 9 1369.0513
10  2835.451
end



gen pc=.
sum ipcm
return list  
replace pc=(ipcm/`r(sum)')*100
drop ipcm
replace pc=round(pc, 0.1)
gen spc =string(pc)

replace spc = subinstr(spc, ".", ",", .)
replace spc = subinstr(spc, "-", "", .)
replace spc=spc+",0" if ~strpos(spc,",")
replace spc ="0" + spc  if (pc>0 & pc<1)
replace spc ="0" + spc  if (pc<0 & pc>-1)
replace spc =spc + "%"

gen var=""
local la "I II III IV V VI VII VIII IX X"
forvalues n = 1/10 {
 local nla: word `n' of `la'
 replace var="`nla'" if d_ipcm==`n'
 }

expand 2  /*duplicar observaciones*/
bysort d_ipcm: gen lgroup = _n
gen r=1
gen str20 s = cond(mi(r),var,var + " " + spc) if lgroup == 1