IgorReader Q calculation needs fixing/improving

[andyfaff]

At the moment the Q calculation in sas/sascalc/dataloader/readers/IgorReader.py uses the small angle approximation for calculating Q. This would produce significant distortions at small detector lengths. It should use arctan instead.

[krzywon]

The pixel size in the reader is also wrong. It should be 5.08 mm, not 5.00 mm. It shows up three times in the reader. Once at the bottom, defined as pixel. The other two times are in the qx and qy calculations, but only as 0.5.

[butler]

I think this actually raises the question once again of whether SasView should continue supporting this format. As I recall we had that discussion a few years ago and there was some suggestion that some people might be depending on it still? and "no harm in keeping it."

I would argue (again:-) that it is time to get rid of this format and these discussions demonstrate exactly why. SasView should not be trying to keep up with the individual geometries of every SAS instrument worldwide. That is the point of "reduced data formats." This format I believe was the first one implemented over 10 years ago before that separation was completely clear and in fact is subsequent discussions are what led to the simple vector format of reduced data until the recently approved NXcanSAS format created a standard which we now also support.

I suggest this would be a good topic to discuss at code camp

[krzywon]

Now that the igor reader has been removed, this ticket is obsolete.