Graphics Reference
In-Depth Information
graph hbox tenure, nooutsides over(occ5, gap(*.2) ) over(collgrad)
Here, we shrink the gap between the
levels of collgrad , making the gaps
20% of the size they normally would.
This yields boxes that are wider than
they normally would.
Uses nlsw.dta & scheme vg past
Prof/Mgmt
Sales
not college grad
Clerical
Labor/Ops
Other
Prof/Mgmt
Sales
college grad
Clerical
Labor/Ops
Other
0
5
10
15
20
Curr. work exper.
excludes outside values
graph hbox tenure, nooutsides over(occ5, gap(*.4) ) over(collgrad, gap(*2) )
We can control the gap with respect to
each of the over() variables. In this
example, we make the gap among the
occ5 categories small (40% of their
original size) and the gap between the
levels of collgrad larger (two times the
normal size).
Uses nlsw.dta & scheme vg past
Prof/Mgmt
Sales
not college grad
Clerical
Labor/Ops
Other
Prof/Mgmt
Sales
college grad
Clerical
Labor/Ops
Other
0
5
10
15
20
Curr. work exper.
excludes outside values
By default, the boxes formed by over() variables are ordered in ascending sequence
according to the values of the over() variable. Stata allows us to control the order of the
boxes by allowing us to put them in descending order, order them according to the values of
another variable, or sort the boxes according to their medians. These options are illustrated
in the following examples.
The electronic form of this topic is solely for direct use at UCLA and only by faculty, students, and staff of UCLA.
Search WWH ::




Custom Search