[RkWard-devel] Length and NAs

Thomas Friedrichsmeier thomas.friedrichsmeier at ruhr-uni-bochum.de
Mon Mar 26 09:35:13 UTC 2007


On Monday 26 March 2007 00:08, Stefan Roediger wrote:
> Fine. I was thinking about to add an feature in the after the coming
> release which allows to use transformed vallues (e.g. "stem(log10(var))").
> But maybe its better to create a plugin which transforms data in a certain
> way or does other operations like summation of some variables to a new one.

Yes, I think this is the way to go, as this would really be nice for virtually 
all plugins. There should also be a way to "inline" such transformation 
plugins, so you can directly use them while filling in the <varslot>. So much 
to do...

> Actually, I really would like to keep it that way because the outlier tests
> are tricky to interpretate sometimes (similar applies to moments). Using
> the descriptive means to run antother test for a rather simple information
> which needs more scrolling and stuff like that (btw, its even more
> convenient to use it that way later on in spread sheet). IMHO, it's the
> fastest way to find out what the data look like.
> I guess these plug-ins won't be extended in future and thus it should be
> quite good maintainable even with this additional "tests". However, you are
> the boos.

Ok, alternative suggestion: Group all those in one option "print 
descriptives". Since they are just in this plugin for quick and easy access, 
I don't think we need this level of fine grained control over what gets 
shown, here. My feeling is, separating all the options, is just more 
checkboxes for the user to check/uncheck, and not really helpful 
configurability.

> Yes, in my opinion they are.

Ok, then I'll take a look, soon.

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/20070326/f3aa3813/attachment.sig>


More information about the Rkward-devel mailing list