Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: RE: Scatterplot matrix

From   Richard Goldstein <>
Subject   Re: st: RE: Scatterplot matrix
Date   Fri, 07 May 2004 08:31:59 -0400

in addition to Nick's comments on graph in version 8,
I think a reminder that one can always use the
version 7 syntax by starting with the command -gr7-
is important.  This is especially true if you are
in an analytic mode (rather than looking for a
publication quality graph) since gr7 is MUCH faster
than the version 8 graphs.


Nick Cox wrote:

. graph matrix ...
Actually, there's a wrinkle here
which I think is worth spelling out.
Presuming that your response variable is y and the others are covariates, the command
. graph matrix x1 x2 x3 x4 y
ensures that the bottom line of graphs all have y as vertical axis variable and the various x* as horizontal axis
variables. You may agree with me in preferring this arrangement to that produced by
. graph matrix y x1 x2 x3 x4
So mention the response _last_ (clearly, the opposite of what you must
do with e.g. -regress-).
When you are unsure of command names,
. search graph matrix
(e.g.) will point towards them.
TELHAJ {Shqiponje|Shqiponja}

After a regress command with -robust- and -cluster- option I used a
Ramsey test (-ovtest, rhs) which suggested that there are higher order
trends, but omitted from the regression model.  I was trying to use a
scatterplot matrix to look at higher order trends.

As I remember in Stata 7, the command for this was:

Graph y x1 x2 x3 x4, matrix

*   For searches and help try:

*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index