You are wrong. The help for -ciplot- indicates
that it takes the options of -scatter-, which
include multiple y axes.
That said, these should surely be identical to each
other for the plot to make any statistical sense.
Nick
[email protected]
Richard Palmer-Jones
> Thanks - ciplot nearly does it - especially getting rid of one set of
> legends (from the twoway rcap and scatter plots).
>
> But I would like a bit more control - to use multiple y axes for
> example which I don't think I can get with ciplot - correct me if I am
> wrong ?
>
> I looked at the code for ciplot, thanks to Nick, and worked out that
> the trick is to use "legend(order(1 2 3)) to drop the redundant
> labels from multiple twoway plots.
Nick Cox
> > I don't think Richard's missing anything
> > w.r.t. -serrbar-. It's been around for
> > a long time, but is not a very general
> > command and does no more than is claimed
> > in the help.
> >
> > In addition to Maarten's suggestion, -ciplot-
> > from SSC may be of use or interest.
> >
> > On the whole, however, CIs plotted just
> > the way you want requires recourse to
> > a combination of -twoway- types, as Richard says.
> >
> > I can't comment on Richard's code because he
> > doesn't show it to us.
Maarten buis
> > > Have a look at Roger Newson's -eclplot- (type -findit eclplot-).
Richard Palmer-Jones
> > > > I would like to overlay means and cis of several variables
> > > against (for
> > > > example) time. Serrbar alows only one variable. twoway line
> > > or scatter
> > > > and rcap can be overlaid for each variable and combined
> in the same
> > > > graph but in the legend I get two entries for each
> varaible and I
> > > > cannot get rid of one of them.
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
