BUG: Groupby.quantile segfaults for quantile > 1

Adding tests for #27467 surfaced some bugs:

from pandas import *
df = DataFrame(dict(a=[0, 0, 0, 1, 1, 1], b=range(6)))
g = df.groupby(np.repeat([0,1], 3))
# ok
g.quantile(50)
<segfault>

quantile accepts floats in [0,1] or arrays of floats. It appears not to validate the range though.

[jbrockmendel]

Apparently this is platform specific. Works fine for me on OSX with numpy 1.16.4, segfaults on Ubuntu 18.04 with numpy 1.16.2

Thanks for investigating