Back to the main page.
Bug 1809 - annoying warning when calling nanvar with w=[]
| Status | CLOSED FIXED |
| Reported | 2012-11-06 13:29:00 +0100 |
| Modified | 2014-01-29 13:28:32 +0100 |
| Product: | FieldTrip |
| Component: | core |
| Version: | unspecified |
| Hardware: | PC |
| Operating System: | Windows |
| Importance: | P3 normal |
| Assigned to: | Eelke Spaak |
| URL: | |
| Tags: | |
| Depends on: | |
| Blocks: | |
| See also: |
Eelke Spaak - 2012-11-06 13:29:29 +0100
line 36: warning('Emulating undocumented behaviour (w=[]). Please use w=0 instead.'); This is OK when the user is calling the nanvar, but ttest() also uses nanvar with w=[]. Therefore, each ttest() generates this warning. 2 possible solutions: - Replace warning() with warning_once(). - Remove the warning, and just implement the behaviour with w=[] as the stats toolbox and don't care that it's undocumented.
Bart Gips - 2012-11-06 15:52:34 +0100
My view: w=0 is the default value, therefor an empty w should always revert to w=0. I would vote for removing the warning. Perhaps we can make a note about this in the function header/help. Unless someone can convince me of a difference in meaning between nanvar(x) and nanvar(x,[]).
Eelke Spaak - 2012-11-07 14:58:48 +0100
as discussed in FT meeting, warning is removed