In 4.2 polydisperse hollow_cylinder @S(Q) fails

[richardh]

In 4.2 hollow_cylinder fails if have both an S(Q) and polydispersity.

File "sas\sascalc\data_util\calcthread.pyc", line 274, in _run
ValueError?: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()

This is one of the two special models, along with vesicle - see also #1257 - where shell not total volume provides the scaling.

In 5.0 there is NOT an extra volume parameter, so this fail does not occur, but we still need to look at the scaling of the model.

[yunliu]

For the scaling issue, a new ticket (#1259) is created.
Even though the scaling issue was only tested for sphere and vesicle with the hard sphere interaction this time, it is useful to check if this is a generic problem for many other models.

[pkienzle]

Cannot reproduce. Please provide parameters that fail.

[richardh]

Replying to pkienzle:

Cannot reproduce. Please provide parameters that fail.

running e.g. release 4.2.1 on windows 10, seen this on two different laptops, used default params for hollow_cylinder, add hard sphere, turn on polydispersity, set pd_radius to 0.1

However my developer build of 4.2 master does not do this, so suspect that either some more recent change has actually already fixed this problem or it is specifically a problem with the release build??