I agree with Giuseppe, stats by default should be calculated in at most 5% of pixels, although there should be a tool to calculate full stats and save them. A random 5% sample is actually a better estimate of the real min,max, because most of the time you do not hit outliers because they a re rare by definition.
Agus El día 28 de febrero de 2012 13:43, Giuseppe Sucameli <[email protected]> escribió: > Hi, > > On Tue, Feb 28, 2012 at 1:07 PM, Alexander Bruy > <[email protected]> wrote: >> Right currently stretching is off by default. But collecting stats >> is necessary not only for stretching. For example some plugins >> needs raster statistics for their work. > > are you talking about approximated stats, isn't it? Because in this moment > the QgsRasterProvider compute full stats only, so the process can require > a lot of time. > > In addition, I think we should re-style the piece of the raster Layer Property > dialog to set min/max (I know there's a ticket yet): > > - the checkboxs are not checked whenever at least one should be, > > - "Estimated (faster)" compute only approximated max/min, instead "Actual > (slower)" compute full stats (why? shouldn't compute min/max either)? > > - the "Load" button is far away from the checkboxes (I thought them was > unrelated). > >> 2012/2/28 Nathan Woodrow <[email protected]>: >>> I thought stretching was off by default... >>> >>> - Nathan >> >> -- >> Alexander Bruy >> _______________________________________________ >> Qgis-developer mailing list >> [email protected] >> http://lists.osgeo.org/mailman/listinfo/qgis-developer > > > > -- > Giuseppe Sucameli > _______________________________________________ > Qgis-developer mailing list > [email protected] > http://lists.osgeo.org/mailman/listinfo/qgis-developer _______________________________________________ Qgis-developer mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/qgis-developer
