Statalist


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

RE: st: RE: RE: -graph twoway- and x-axis positioning


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: RE: RE: -graph twoway- and x-axis positioning
Date   Tue, 18 Nov 2008 22:01:29 -0000

I don't think anyone was claiming otherwise. But rightly or wrongly,
StataCorp by making choices about what is easy or available create
consequences for what is difficult or impossible. For example, StataCorp
do not explicitly support dotted or striped or similar textures, and
although, as Sergei showed not so long ago, that is not an absolute
barrier, it sure makes things elusive for non-Sergeis. And, in that
case, my guess is that enthusiasts will wait indefinitely for such
features. 

The serious point is that Stata graphics is not really a low-level
graphics language in which you can easily start with a tabula rasa and
add precisely what you wish. It is a high-level graphics language with a
high (but not indefinite) degree of low-level control. 

Equally, I don't think StataCorp wish for a instant to deny individual
users their right to strong personal preferences (prejudices, if anyone
prefers) about graphics, as expressed for example in this thread.

Nick 
n.j.cox@durham.ac.uk 

Jeph Herrin

I think StataCorp in general should not decide what graphs
people _should_ make, and opt instead for maximizing what
they _could_ make.

I much prefer to do all of my graphs in Stata for all of
the reasons Scott gives, but too often (ie, at least once)
have found myself dumping the data to Excel to add or remove
some feature as determined by colleagues or (more often)
journal editors.


Nick Cox wrote:

> So, it's a virtue if Stata can reproduce a lousy graph easy in Excel?
> Don't answer that one. 
> 
> Scott has meanwhile gone most of the way. His ticks could be made into
> little rspikes. 
 
Stephen P. Jenkins
 
> The answer to Nick and Austin's opening questions is "yes" -- though I
> strongly agree with their additional remarks about the general
> undesirability of doing this (as my final sentence hinted).
> 
> The background: I am trying to persuade some sceptical colleagues to
> change from using Excel for graphics to using Stata for graphics --
> for reproducibility and audit trails, in addition to functionality and
> 'total control'.  I have been given a series of Excel graphs as
> 'challenges' by these colleagues, including one with this axis
> feature.  This is the only thing that has beaten me so far.
> 
> Nick Cox 
> 
> Do you mean you want ticks and x-axis labels superimposed within the
> data region? I'd refuse to play, regardless of whether it can be done.
> 
> Austin Nichols 
> 
> You want the ticks and labels moved into the inner plot region??
> Seems like a bad idea, but you could write a little pre-processing
> program to accumulate the necessary ticks (lines) and labels (text) in
> a local.  Is it worth the effort?

*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index