st: RE: RE: RE: sequential graphing

[Nick Cox <n.j.cox@durham.ac.uk>]

I don't have any deep thoughts on this. What was said about markers applies to line elements too.  If there are lots of occlusions, and the result is to you an unacceptable mess, then you need a different design. No small tricks will help enough. 

-by()- with a -total- option is a Stata standard. 

As Bill Gould once emphasised to me, many graphs can be made small without enormous loss. As he might have said, a messy graph shown big is still a messy graph. 

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

Lim, Raymond

I guess a more concrete example of pcspike would clarify. This is a completely random example. Let's say you have a map of Central Park. You have beginning x-y coordinate of a trip and end x-y. You connect these points with a red line for trips that are completed earlier than expected or on time, and a blue line for trips that are completed later than expected. In my graphing command, if I specify the red first and blue next; the blue lines washes out the red. Vice versa if I swap the order. 

The code would be something along the lines of:
twoway pcspike y_start x_start y_end x_end if early==1, lwidth(vvthin) lcolor(red)  ///
|| pcspike y_start x_start y_end x_end if early==0, lwidth(vvthin) lcolor(blue) 

Again, this is a completely random example. 

*
*   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/