Create to_field function

[ericspod]

Create a function called to_field which accepts something that can be coerced into a Numpy array and returns a new field containing that data. Other arguments such as dtype may be needed. The method should use np.asarray to coerce the input to an array. The dtype of the array should be checked to see if it's a known type, if not raise an exception.

For example:

n=np.random.rand(10,dtype=np.float32)

f=exetera.core.fields.to_field([1,2,3])  # using asarray this can be used to create a field from Python array data

f=exetera.core.fields.to_field(n)  # returns a new field containing float32

f1= exetera.core.fields.to_field(n) + n  #  allows adding numpy arrays to the new field

Do we want a to_categorical_field as well?