Using hv.Vectorfield: AttributeError: 'float' object has no attribute 'cos'

[ahuang11]

import numpy as np
import holoviews as hv
hv.extension('bokeh')
hv.VectorField(([-1, 0, 1, 2, 3], np.deg2rad([0, 45, 90, 180, 360])))

[ahuang11]

from @ea42gh

>>> h=hv.VectorField(([-1, 0, 1, 2, 3], np.deg2rad([0, 45, 90, 180, 360])))
>>> h.data
   x         y Angle Magnitude
0 -1  0.000000   NaN       NaN
1  0  0.785398   NaN       NaN
2  1  1.570796   NaN       NaN
3  2  3.141593   NaN       NaN
4  3  6.283185   NaN       NaN

This works:
hv.VectorField(([-1, 0, 1, 2, 3], [-1, 0, 1, 2, 3], np.deg2rad([0, 45, 90, 180, 360])))

So maybe a better error for not enough dimensions?

[philippjfr]

Definitely down to the datatype and missing nans:

np.cos(np.array([float('nan')], dtype='object'))

The solution here is that to ensure that the pandas interface will not let you create an vectorfield without all the dimensions.