[Qgis-developer] A bug in raster stats?

[Benoit de Cabissole]

Hi Manuel,

Thanks for the clarification, makes sense!

Cheers,
Benoit


> -----Original Message-----
> From: Manuel Massing [mailto:m.massing at warped-space.de]
> Sent: Thursday, 12 August 2010 10:47
> To: qgis-developer at lists.osgeo.org; benoit at exigesa.com
> Subject: Re: [Qgis-developer] A bug in raster stats?
> 
> 
> Hi Benoit,
> 
> > Sum of squares
> > 40194.0016233767
> > Sum of all cells
> > 8931.0000000000
> 
> The stats show the empirical variance of the data (i.e. sum 
> (value-mean)^2 ),
> whereas 169078 is the sum of uncentered squares.
> 
> Maybe the property could be renamed to "sum of squared 
> deviations" to clarify 
> this distinction?
> 
> As for the sum, that was  an off by one error (fixed in r14069). 
> Good catch :-)
> 
> bye,
> 
> 	Manuel