Re: [ePiX-users] Error in histogram documentation

[Marcus D. Hanwell]

Hi Andrew,

On Tuesday 02 January 2007 18:04, Andrew D. Hwang wrote:
> On Tue, 2 Jan 2007, Marcus D. Hanwell wrote:
> > I think there is an error in the documentation and it nearly made me give
> > up on ePiX for plotting histograms. The documentation lists n (the number
> > of bins) as being the third optional argument. This appears to have been
> > changed (found the entry in the ChangeLog too) but the documentation has
> > not been updated. It is now the first optional argument.
>
> I'll have a careful look at data plotting today, and will consider adding
> alternative plotting and axis labeling styles (log, trig, fractions, ...).
> At the very least I'll correct and expand the manual, but please feel free
> to pass along any additional suggestions for data plotting interface and
> features.

I will happily test these for you - I am hoping to use ePiX to plot all of my 
data if possible as I really like the interface and the graphs look great in 
LaTeX documents. I program a lot in C++ so it makes sense to do my analysis 
using ePiX programs if possible.
>
> Non-Cartesian axis and (2-D) plotting styles were in the back of my mind
> years ago, and it's probably time to think about how things "ought" to
> work. At risk of introducing more global data, one obvious approach would
> attach a flag to the current drawing state (cartesian, polar, log-log,
> log-lin, etc.), and have the state controlled similarly to angular units,
> color, or line width. Since an input file corresponds fairly tightly to a
> single picture, global data in epix-1 doesn't bother me as much as it
> Should. In epix-2, plot style will be safely encapsulated in appropriate
> objects. I could do the same for epix-1, but that's not a small change,
> and leads to the next thought:
>
> Several aspects of epix's internal behavior persist because they work, if
> clumsily. However, I'm now starting to consider the long-term benefits of
> rewriting basic chunks of ancient, kludgey code. If this happens, I'll
> move this thread to the devel list, and will revert "Version 1.1rc1" to
> "Version 1.0.22". :)
>
> Thanks for raising these issues, and for reporting the manual errors!
>
It is OK, thanks for getting back to me so quickly. I would be more than happy 
to test any updates. I was looking at the ePiX 2 stuff but I wasn't sure if 
that was ready to be used yet. I can work with CVS/SVN versions if that makes 
things easier - if useful I could also provide diffs if I find bugs I can fix 
in a useful way.

I would really appreciate the log-lin axes and different labelling styles. If 
it is helpful I could probably dig through the code and help to patch it up 
if you don't have time. I didn't want to end up maintaining a personal patch 
set others might also find useful.

Thanks,

Marcus