avoid errors when encountering arrivals without time residual set

[megies]

just ignore them during plotting..

See #49, @ChrisBail let me know if this works for you..

[ChrisBail]

Won't it be better to plot them (dashed or gray) instead of totally ignoring them (just wondering)? Anyway, If you feel it's better this way then keep it that way

[megies]

Well, I think plotting them with the same time as the pick is misleading, because then you would have to assume that the residual is 0.0 (i.e. "pick fits perfectly with best hypocenter + velocity model used") when it's rather not even set (i.e. "time residual of pick is unknown").

Honestly, I don't even know what to make of an arrival with unset residual time. I'm interpreting this as 'the pick was correctly assumed to be phase XYZ as described by pick.phase_hint but location routine can't give a residual time'.