Hi! When I try to work with expressions like (-X)^{1/2} or any other fractional degree the expand_power and collect_powers commands cause a kernel crash, for example:

`collect_factors($(-X)**{-5/2} X$);`

I know that without defining the X sign beforehand, I can't shorten this expression, but kernel crush is too much. That's why I have another question for this bugreport, is it possible to define the sign of X in advance to work with such expressions correctly?