Statalist


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

RE: st: AW: AW: Problematic twoway lineplot with -by-


From   "Nick Cox" <n.j.cox@durham.ac.uk>
To   <statalist@hsphsun2.harvard.edu>
Subject   RE: st: AW: AW: Problematic twoway lineplot with -by-
Date   Thu, 13 Aug 2009 18:20:21 +0100

If you like -separate- check out 

SJ-5-4  gr0023  . . . . Stata tip 27: Classifying data points on scatter plots
        . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  N. J. Cox
        Q4/05   SJ 5(4):604--606                                 (no commands)
        tips for using separate, gray-scale gradation, and text
        characters as class symbols to classify data points on
        a scatter plot

Accessible to all at 
http://www.stata-journal.com/sjpdf.html?articlenum=gr0023

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

Steinar Fossedal

Thank you, Martin!

-separate- was quite the nugget for my repetoire, it will help a lot
in the future.

2009/8/13 Martin Weiss <martin.weiss1@gmx.de>:

> Last two lines should be
>
>
> *************
> gr combine graph1 graph2, /*
> */ rows(1)
> *************
>
> These things do not always transfer well...

Martin Weiss

> *************
> // Create dataset
> clear*
> set obs 40
> gen id = int((_n-1)/10)+1
> gen group = (id>=3) + 1
> bysort id: gen time = _n
> replace time = time -.5 if id  == 4
> gen value = uniform()
> bysort id (time): replace value = /*
>  */ value[_n-1] + uniform() if _n > 1
>
>
> separate value if group==1, by(id)
>
> tw (line value1 time if group==1, sort) /*
>  */ (line value time if group==2, sort), /*
>  */ legend(off) nodraw
>
> graph rename graph1
>
> tw (line value2 time if group==1, sort) /*
>  */ (line value time if group==2, sort), /*
>  */ legend(off) nodraw
>
> graph rename graph2
>
> gr combine graph1 graph2,
> rows(1)
> *************

Steinar Fossedal

> I'm trying to create a twoway graph using two line graphs, and have it
> split into groups using -by- _on_only_one_of_the_graphs_. I'm finding
> this a bit tricky to pull off, and I've cooked up an example below
> which illustrates the problem. My goal is to have graph 2 included on
> each of the plots in graph 1, resulting in a graph with two plots
> having two lines each. Instead, I get a graph with four plots in it,
> and only a single line per plot.
>
> Is there any way to produce the graph that I want without rearranging my data?
>
> Example :
>
> // Create dataset
> clear
> set obs 40
> gen id = int((_n-1)/10)+1
> gen group = (id>=3) + 1
> bysort id: gen time = _n
> replace time = time -.5 if id  == 4
> gen value = uniform()
> bysort id (time): replace value = value[_n-1] + uniform() if _n > 1
>
> // Create plots
> // Graph 1 (using by)
> twoway (line value time if group==1, sort by(id))
> graph rename graph1
> // Graph 2 (not using by)
> twoway (line value time if group==2, sort)
> graph rename graph2
> // Graph 3 (Attempt to merge the two graphs together, getting graph 2
> on each of the plots in graph 1)
> twoway (line value time if group==1, sort by(id)) (line value time if
> group==2, sort)
> graph rename graph3

*
*   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