inadequate variability Boxplot

24 views
Skip to first unread message

Vassilios Mazis

unread,
May 13, 2016, 9:08:24 AM5/13/16
to sofastatistics
Hello, 

first of all, thanks. Usability Is a big plus ! 

I have come up with a real set of data that refused to Boxplot a subgroup due to "inadequate variability".  I just tweaked the response, until it did, so I know that it was not the number of samples that was inadequate. 
However, this has the effect of leaving my audience with the impression that no data were collected. 

I wonder (no theoretical knowledge here) if it would be preferable to show the mean and force the whiskers to collapse or disappear ? 


once again, thanks


VM 

Grant Paton-Simpson

unread,
May 15, 2016, 4:25:49 PM5/15/16
to sofasta...@googlegroups.com
Hi Vassilios,

It would be easy for me to make this change for boxplots for two variables (one averaged and one grouping) but I'll have to think a little harder to make it work for boxplots with 3 variables (one averaged, one within the series, and one grouping separate series). Almost certainly can do it but best to do it when I get a good crack at it in one go. Thanks for raising the issue. Probably be part of next release.

All the best,
Grant
--

---
You received this message because you are subscribed to the Google Groups "sofastatistics" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sofastatistic...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Grant Paton-Simpson

unread,
May 15, 2016, 5:04:38 PM5/15/16
to sofasta...@googlegroups.com
Hi Vassilios,

I have managed to make that work (and added some more comments in the code to make it easier to work with the underlying boxplot functions) - are you interested in testing the pre-release version? If so, are you on Windows, Mac, or Linux?


All the best,
Grant

On 14/05/16 01:08, Vassilios Mazis wrote:

Vassilios Mazis

unread,
May 16, 2016, 9:26:46 AM5/16/16
to sofastatistics
Hey, thanks ! 

I am on windows. and can probably talk a sysadmin into doing linux too. 

thanks for being so responsive. caution ! I may ask for more ! 


VM 

Vassilios Mazis

unread,
May 20, 2016, 5:58:08 AM5/20/16
to sofastatistics
Hello again,

I can confirm that repeated or not-so-different values no longer prevent a box result whilst charting my datasets. 
nice. 

many thanks from Greece  

VM 


PS. however, during testing I have noticed something surprising with filter behaviour, in this and the previous version, so I may come back on this ! 


 


Grant Paton-Simpson

unread,
May 20, 2016, 6:12:48 AM5/20/16
to sofasta...@googlegroups.com


On 20/05/16 21:58, Vassilios Mazis wrote:
Hello again,

I can confirm that repeated or not-so-different values no longer prevent a box result whilst charting my datasets. 
nice.
Phew


many thanks from Greece  

VM 


PS. however, during testing I have noticed something surprising with filter behaviour, in this and the previous version, so I may come back on this !
Please do - the fewer surprises in our software (apart from ground-breaking analyses) the better.

Vassilios Mazis

unread,
May 20, 2016, 6:31:39 AM5/20/16
to sofastatistics
done as a PM, to avoid a false alarm ! 
Reply all
Reply to author
Forward
0 new messages