How to 'zoom' in on the central area of a graph, after mca

[Niamh Caffrey]

I am using Stata 15.1
I would like to create a secondary graph that focuses in on the output of a joint correspondence analysis. The sample code below uses the dataset as per the Stata manual info for MCA. When this graph is created i would ideally like to be able to place a rectangle around the points x = -2,2 y = -2,2 and then produce another graph that focuses on that area. Can someone please help me with code to do this?

[CODE][
use http://www.stata-press.com/data/r15/issp93
describe

mca A-D, method(joint) normalize(principal)
mcaplot, overlay origin
/CODE]

[Andrew Musau]

Thanks for the reproducible example. I do not think that you will be able to do this within mcaplot. Your best bet is to reconstruct the graph using twoway. I suspect that there is a way to get the inputs used to create the mcaplot graph using post estimation commands, but I am no expert here. However, Scott Merryman shows in this link how you can recover the inputs from any Stata graph, so here is a first attempt at recreating the graph.

use http://www.stata-press.com/data/r15/issp93, clear
describe
mca A-D, method(joint) normalize(principal)
mcaplot, overlay origin
graph save "Graph1", replace
clear
gr use "C:\Users\709554\Desktop\Graph.gph"
serset dir
serset use
local names ""
foreach l in A B C D{
     local names "`names' dimension1`l' dimension2`l' coord`l'"
}
rename ( __000000- __00000B ) (`names')
gen id=_n
reshape long dimension1 dimension2 coord, i(id) j(which) string
tw (scatter dimension1 dimension2 if which=="A", mlab(coord)) ///
(scatter dimension1 dimension2 if which=="B", ///
mlab(coord)) (scatter dimension1 dimension2 if which=="C", ///
mlab(coord)) (scatter dimension1 dimension2 if which=="D", ///
mlab(coord) xline(0) yline(0)  title("MCA coordinate plot"))
graph save "Graph2", replace
gr combine Graph1.gph Graph2.gph

You can now add your manipulations. You can restrict the data using the -if- qualifier.