Le 12/08/2012 22:46, Nathaniel Echols a écrit :

On Sun, Aug 12, 2012 at 12:48 PM, Pascal wrote:

...

Is cctbx doing all the calculations in double or single precision? I am intersted in the direct method of the structure factors calculation. From what I saw it's double precision but I am not sure.

As far as I know everything is double-precision in practice - the majority of the C++ code is written as templates so in theory could be either single- or double-precision, but the latter is always what gets used.

Thanks, Pascal