InequalityLoss

[rdrivers]

If InequalityLoss is called with float arguments for lower and upper, line 28 of losses.py raises an assertion error, since it seems to be expecting numpy arrays, not scalars. It would be nice to have an example or documentation on the use of this, though it produced the correct answer when I used InequalityLoss(.15, np.array([-.05]), np.array([.05])), constraining the results to be between .1 and .2.