REGR: np.i0(Series) failing with IndexError on NumPy 1.17.0rc1

[TomAugspurger]

In [2]: s = pd.Series([1, 2, 3])

In [3]: np.i0(s)
---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
<ipython-input-3-0fe1635dc55e> in <module>
----> 1 np.i0(s)

<__array_function__ internals> in i0(*args, **kwargs)

~/Envs/dask-dev/lib/python3.7/site-packages/numpy/lib/function_base.py in i0(x)
   3104     """
   3105     x = np.abs(x)
-> 3106     return piecewise(x, [x <= 8.0], [_i0_1, _i0_2])
   3107
   3108 ## End of cephes code for i0

<__array_function__ internals> in piecewise(*args, **kwargs)

~/Envs/dask-dev/lib/python3.7/site-packages/numpy/lib/function_base.py in piecewise(x, condlist, funclist, *args, **kw)
    621             y[condlist[k]] = item
    622         else:
--> 623             vals = x[condlist[k]]
    624             if vals.size > 0:
    625                 y[condlist[k]] = item(vals, *args, **kw)

IndexError: too many indices for array

[TomAugspurger]

I'm not sure what would have broken this.

And... it may be numpy 1.17 only Doing some more debugging.

[TomAugspurger]

Yeah, it's just with NumPy dev

In [2]: s = pd.Series([1, 2, 3])

In [3]: np.__version__
Out[3]: '1.16.0'

In [4]: np.i0(s)
Out[4]: array([1, 2, 4])

In [1]: import numpy as np; import pandas as pd

In [2]: s = pd.Series([1, 2, 3])

In [3]: np.i0(s)
---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
<ipython-input-3-0fe1635dc55e> in <module>
----> 1 np.i0(s)

<__array_function__ internals> in i0(*args, **kwargs)

~/Envs/pandas-dev/lib/python3.7/site-packages/numpy/lib/function_base.py in i0(x)
   3104     """
   3105     x = np.abs(x)
-> 3106     return piecewise(x, [x <= 8.0], [_i0_1, _i0_2])
   3107
   3108 ## End of cephes code for i0

<__array_function__ internals> in piecewise(*args, **kwargs)

~/Envs/pandas-dev/lib/python3.7/site-packages/numpy/lib/function_base.py in piecewise(x, condlist, funclist, *args, **kw)
    621             y[condlist[k]] = item
    622         else:
--> 623             vals = x[condlist[k]]
    624             if vals.size > 0:
    625                 y[condlist[k]] = item(vals, *args, **kw)

IndexError: too many indices for array

In [4]: np.__version__
Out[4]: '1.17.0rc1'

[TomAugspurger]

Seeing if this is a NumPy issue numpy/numpy#13894

[TomAugspurger]

Fixed upstream by @seberg. Thanks!