re: st: easy histogram

[David Airey <david.airey@vanderbilt.edu>]

> David makes a good point for removing the option for -histogram-, and 
> we will
> remove the checkboxes for log scales from the easy graph dialog for
> -histogram-.
>
> However, the "full featured" dialog for -histogram- will remain the 
> same since
> both -xscale(log)- and -yscale(log)- are valid -graph twoway- options.
>
> --Jeff
> jpitblado@stata.com
This reply beats around the bush and doesn't explain to me why 
"histogram mpg, xscale(log)" would ever make sense. That is really my 
question; I'm ignorant of the answer.

Jeff's answer that histogram's option xscale is valid because it is a 
twoway explains why histogram inherits these options. That doesn't mean 
the design choice is a good one. The area under the histogram should 
sum to 1 when the yaxis is density, as stated in the manual. I don't 
think when xscale(log) is used it will (though I have not measured). If 
it doesn't, then the option is pointless (the option is not pointless 
for other graph twoway commands like scatter).

Perhaps not all daughters of the mother twoway should inherit certain 
twoway options?

*
*   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/