As you have rightly noted, a long cylinder pointing end on to the radiation source (probably the sun, especially during periods of high activity), can be shielded by a smaller, and thus lighter, shield. On the other hand if you need shielding facing all directions, then a sphere is a better choice.
The spaceship is moving fast, in some direction given by where it is, where it wants to be, and orbital mechanics. Dust, rocks and debris will be moving somewhat at random, but will tend to have a preferred direction. They mostly go round the sun/planet/moon the same way the rest of the solar system rotates, and there won't be many rocks on orbits which pass through another body. So in some cases, there will be a direction from which more rocks come than others. It makes sense to point one end of the long cylinder in that direction to reduce the chance of being hit. Note that the ISS is constructed/orientated with this at least partly borne in mind.
This isn't something which drives the orientation as much as something which might make it unsafe to vary from the orientation the designers planned. You probably have solar panels which want to face the sun, and radiators which don't. If you don't keep an orientation which allows them to do so, and you don't have alternate energy sources/heat sinks, you're going to have problems. Similarly, if the design was expected to face the sun and ends up spinning, the rapid heating/cooling cycles might do bad things to the skin. If it was intended to spin and is now static, thermal management could be difficult.
A 3D object spinning in space is stable when spinning in some directions, but not others. Don't be this shape. It would be wise to pick a rotation which is stable, even if stuff moves around inside the ship. Otherwise you'll be wasting propellant or electricity (on reaction wheels/gyros) staying stable, and have additional risks if that subsystem fails. This isn't about the external shape as much as it is about the distribution of mass inside that shape.
Since you ask about a large ship, and are concerned about radiation, I wonder if you are thinking of a future manned ship. If so, you'll want to set the spin up to give reasonable gravity, uniform gravity within the crew compartment, and minimal Coriolis effect. That way your astronauts are less likely to misjudge it and get hurt, or spend too much time throwing up. Overall safety is going to depend on them making good decisions.
There are probably other considerations I haven't though of...