Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

st: RE: RE: Missing x-axis labels when using graph twoway


From   "Garth Rauscher" <garthr@uic.edu>
To   <statalist@hsphsun2.harvard.edu>
Subject   st: RE: RE: Missing x-axis labels when using graph twoway
Date   Mon, 20 Aug 2007 15:38:17 -0500

Thank you Nick for answering my two questions about graph labeling and sort
order.  

Garth Rauscher
Assistant Professor
SPH Epidemiology and Biostatistics
University of Illinois at Chicago
1603 W Taylor St (M/C/ 923)
Chicago, IL 60612
(312) 413-4317 phone
(312) 996-0064 fax
garthr@uic.edu email 


-----Original Message-----
From: owner-statalist@hsphsun2.harvard.edu
[mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Nick Cox
Sent: Thursday, August 16, 2007 12:28 PM
To: statalist@hsphsun2.harvard.edu
Subject: st: RE: Missing x-axis labels when using graph twoway

#1 

Value labels are not shown by default. You must specify (e.g.) -xla(,
valuelabel)-, as is documented in the help for axis_label_options. 

#2 

Consider 

. sysuse auto
. scatter mpg weight
. sort mpg
. scatter mpg weight 

The -sort- affects the sort order of the data, but is irrelevant to the
scatter plot, which is unchanged. What you asked for "did work"; the problem
is in your expectations. You can't do a scatter plot versus weight using the
sort order of mpg because it would no longer be a scatter plot versus
weight. (You can connect points in that order, but that is a different
story.) 

In a simpler analogue of your data 

users_meeting_venue          order_in_2007

Essen                        1
Boston                       2
Stockholm                    3 
London                       4
Roma                         5 
Los Angeles                  6 

If you -encode users_meeting_venue- you get  

encoded users_meeting_venue          order_in_2007

value 1 label Boston                       2
value 2 label Essen                        1
value 3 label London                       4
value 4 label Los Angeles                  6
value 5 label Roma                         5 
value 6 label Stockholm                    3 

But really what you want to do is to attach the names as labels to
-order_in_2007-! 

One solution to your problem to the -egen- function
-axis()- in -egenmore- from SSC. You create a new variable like this: 

egen x_axis = axis(_se study), label(study) 

and then use -x_axis- as your x axis variable, not -study-. 

It may well be that 

egen x_axis = axis(_se), label(study) 

is enough, but what I wrote first includes protection against ties on
standard error. 

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

Garth Rauscher
 
> I have two probably simple questions about graphing in Stata. 
> #1 involves
> getting value labels to appear on the x axis. #2 involves sorting the 
> graph on a variable other than the x-axis variable.
> 
> 
> #1
> I have a dataset where each observation contains a string variable
> ("study_") and a point estimate and lower and upper confidence limits 
> associated with a single study. I want to plot these estimates and 
> confidence intervals, with magnitude of each estimate on the y-axis 
> and study information on the x axis. Apparently string variables are 
> not allowed with graph twoway, so I encoded a numeric variable (named
> study) from the
> original string variable (named study_), containing the original 
> strings as value labels.
>  
> 
> encode study_, generate(study)
> 
> 
> This appears to have worked just as expected:
> 
> groups study study_
> 
>                      study                         study _    
>       Freq.
> Percent |
>  
> |-------------------------------------------------------------
> --------------
> ------|
>   |         BROWN (1992) WOM(0.2)           BROWN (1992) 
> WOM(0.2)       1
> 5.26 |
>   |        CAPLAN (2003) TOTAL(2)          CAPLAN (2003) 
> TOTAL(2)       2
> 10.53 |
>   |        DEGNAN (1992) TOTAL(2)          DEGNAN (1992) 
> TOTAL(2)       1
> 5.26 |
> 
> 
> 
> groups study study_, nolabel
> 
>     study                 study _           Freq.   Percent |
>   |---------------------------------------------------------|
>   |     1           BROWN (1992) WOM(0.2)       1      5.26 |
>   |     2          CAPLAN (2003) TOTAL(2)       2     10.53 |
>   |     3          DEGNAN (1992) TOTAL(2)       1      5.26 |
> 
> 
> 
> However, when I go to graph these estimates using the numeric variable 
> study, the value labels for study do not appear on the x-axis.
> 
> scatter _se study || rcap hi_se lo_se study
> 
> 
> #2 I'd like to sort the graph in order of ascending standard error of 
> the estimate.
> 
> Neither.....
> 
> sort _se
> scatter _se study || rcap hi_se lo_se study
> 
> Nor....
> 
> scatter _se study, sort(_se) || rcap hi_se lo_se study
> 
> Nor....
> 
> scatter _se study, sort(_se) || rcap hi_se lo_se study, sort(_se)
> 
> Will work.
> 

*
*   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/

*
*   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/



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