[RkWard-devel] Distribution CLT: revision 1417

Thomas Friedrichsmeier thomas.friedrichsmeier at ruhr-uni-bochum.de
Tue Feb 20 13:38:58 UTC 2007


Hi,

On Tuesday 20 February 2007 07:33, Prasenjit Kapat wrote:
> As I had said, these are not probably the most useful ones, but nonetheless
> nice (at least I feel). The plan is to use either the
> histogram/qqplot/distribution function to visualize how a distribution
> approaches normality under CLT assumtions. Only, histogram has been
> implemented yet. No help files either. I hope to improve on this quickly.

this certainly looks pretty nice. Looking forward to see more.

I like the histogram options you provide. Maybe it would make sense to move 
the options for the normal curve to a separate tab, since
a) this will also be relevant for the distribution and qqplots, hence is 
independent of the histogram
b) Maybe we can make the histogram(-options) an <embed>-able plugin. It seems 
much more complete than the Plots->Histogram plugin already, so it would be 
nice, if we could reuse it there as well (it does not have the integer-breaks 
method, yet, though, which I think is pretty useful in other use-cases). 
Also, this will be needed in all CLT plugins, so it would be nice, if you 
wouldn't have to copy it to each single one.

Interestingly, for these plugins, I sometimes see the problem about 
the "Processing. Please wait" being shown continuously, like has been 
reported, recently. I'll have to investigate, what exactly may be going 
wrong, here (but no time to look at it, today).

One issue I noted within the plugin: I think the breaks-parameter needs to be 
specified in the call to hist(), not the plot().

I'll take a closer look, once I have the time, or there is more to see. Keep 
up the good work!

Regards
Thomas
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://mail.kde.org/pipermail/rkward-devel/attachments/20070220/549cdcc5/attachment.sig>


More information about the Rkward-devel mailing list