{smcl}
{* 21jun2004}{...}
{hline}
{center:{hi:Twoway with data manipulation can be flexible}}
{center:{hi:{it:(and isn't always 2 ways)}}}
{hline}

o Hand drawn box and histogram plots {right:{stata graph use boxhist2:use}         }

o Parallel coordinates {right:{stata graph use parp_v:use}         }

o Distribution illustrations {right:{stata do norm_p95:do}  {view norm_p95.do:do-file}}

o Regression illustration {right:{stata do regline_ci:do}  {view regline_ci.do:do-file}}

o Population pyramid {right:{stata do pop_pyr:do}  {view pop_pyr.do:do-file}}

o {cmd:twoway}? {right:{stata graph use triplot_us:use}         }

{hline}
{p 4 4 4}{it}
Note:  click on the {stata *:use} or {stata *:do} links to see the graphs,
click on the {stata *:do-file} links to see the {cmd:graph} commands.{p_end}

{p 4 4 4}{it}
The "box and histogram" was created by Nick Cox using his {cmd:tabplot}
program.  The "parallel coordinates" plot was created with Nick Cox's
{cmd:parplot} program.  The "{cmd:twoway}?" plot was created with Nick Cox's
{cmd:triplot} program.  All three of Nick's programs are available from the
SSC archives.  Just type {cmd:findit triplot} (or one of the other program
names) to locate the program.  The remaining graphs are "simple" {cmd:twoway}
commands which can be seen by clicking {stata *:do-file}{p_end}

{p 4 4 4}{it}
With all of this flexibility and legitimate concerns, why open Pandora's box
and discuss the internals of graphics now?{p_end}

{p 6 9 4}{it}
First, I wanted to talk about something new and, as we can see, many are
already discovering the flexibility of {cmd:twoway}.  In fact, they have
surprised me and taught me new tricks.{p_end}

{p 6 9 4}{it}
Second, we believe that things are stable enough that the things I am going to
show you will continue to work, even though they will remain undocumented.

{p 6 9 4}{it}
Third, and most important, these facilities are cool; and I wanted to talk
about them before another developer beat me to the punch.

{p 4 4 4}{it}
I do not expect everyone to immediately start developing new graphs with the
information from this talk.  We will at best scratch the surface of just a few
aspects of the graphics facilities.  I do expect a few people to surprise me by
using these facilities in unexpected ways.  I also expect everyone will learn
a few useful tricks and gain a better understanding of the structure of Stata
graphics.

{p 4 4 4}{it}
If all of this sounds beyond what you want to hear about graphics, I recommend
beginning with the following resources.{sf}

    {view new_res.smcl:>>Resources for those less familiar with {cmd:graph}} (ancillary)

{hline}
{center:{view lions.smcl:<<}   {view tindex.smcl:index}   {view what.smcl:>>}}