Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: RE: RE: Graph: Control plot region- contract, not expand


From   Amanda Blakely Delp <[email protected]>
To   [email protected]
Subject   Re: st: RE: RE: Graph: Control plot region- contract, not expand
Date   Thu, 16 Jun 2011 12:35:53 -0400

Thanks, James, but what you're suggesting oversimplifies the problem
and eliminates part of the dataset I want to include. Eric properly
expresses it--

I have a dataset for which a < b < c < d of my x-axis variable. I want
to estimate and fit a line over [a, d], but only want the graph to
display [b, c].

This is the base command I'm working with, if that tells you anything:
twoway (scatter var1 var2 if year >= 1990 & year <= 2004 & usegr2==1,
msymbol(smcircle)) (lfit var1 var2 if year >= 1990 & year <= 2004) if
usegr==1, yscale(range(0 40)) xscale(range(0 10))

Is this possible?

Thanks again,
Amanda

On Thu, Jun 16, 2011 at 12:10 PM, DE SOUZA Eric
<[email protected]> wrote:
> I don't think this is what Amanda wanted.  The way I understand her, she wishes to estimate and fit a line over the range (a b) but only show the graph for that fit over  the range (c d) where a < c and d < b
>
>
> Eric de Souza
> College of Europe
> Brugge (Bruges), Belgium
> http://www.coleurope.eu
>
>
> -----Original Message-----
> From: [email protected] [mailto:[email protected]] On Behalf Of Fiedler, James (JSC-SK)[USRA]
> Sent: 16 June 2011 17:55
> To: [email protected]
> Subject: st: RE: Graph: Control plot region- contract, not expand
>
> Hello,
>
> I haven't seen a response to this yet, so I'll take a stab at it.
>
> I'll have to make some assumptions, since you haven't explicitly said which command you are using.
>
> I assume you're using -twoway lfit-.
>
> Here's another option, which forgoes the convenience of -twoway lfit- but gives extra control:
>
> regress y x
> predict yhat
> twoway (scatter y x if x >= 1 & x <= 10) (line yhat x if x >= 1 & x <= 10, sort(x))
>
>
> James
>
> ________________________________________
> From: [email protected] [[email protected]] On Behalf Of Amanda Blakely Delp [[email protected]]
> Sent: Thursday, June 16, 2011 10:13 AM
> To: [email protected]
> Subject: st: Graph: Control plot region- contract, not expand
>
> Hello all,
>
> I'm trying to create a fairly simple graph- just a scatterplot overlayed with a linear prediction.  My problem is that I can't seem to control which area of the graph to display. I know I can expand the plot region, but there doesn't seem to be a similar code to force-contract it.
>
> All I care about is the 1-10 region on my scale (x-axis), where some outliers are outside of this region.  Of course, if I drop these outliers, the graph region cooperates.  Is it possible to crop the display region without dropping the observations which determine the behavior of the linear prediction?  Like a way to force the command
> "xscale(range(0 10)"?
>
> Ideally, I'd like to compare 2 graphs-- one with the outliers, and one without, but only along the smaller region of the x-axis.
>
> Many thanks-
> Amanda
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index