[AstroPy] reproject and flux conservation

[Yannick Roehlly]

Hi all,

I'm working with a student who needs to have a visible image rebinned to the 
pixel size of an infra-red one with one to one pixel overlap. I suggested it 
would be simpler to use `reproject` to re-project the visible map onto the 
infrared one.

What we don't understand, is that whereas reproject_exact is supposed to 
conserve the flux, the sum of the reprojected array is much higher that the 
sum of the initial one (I would have expected it to be lower because of parts 
of the visible map not being on the IR coverage).

Where are we wrong?

Regards,

Yannick

-- 
Every man who is high up likes to think that he has done it all himself,
and the wife smiles and lets it go at that.
		-- Barrie