fixed issue with single ensemble concatenation by allowing for choice of which axis to add to 1D arrays

[jscora]

Problem & Solution Description (including issue #)

This fixes an issue I ran into that concatenating qp ensembles with 1D ancillary data didn't work.

• I created an ensure_2d_array function in dictionary.py to replace np.atleast_2d use in concatenate_dicts
• new function allows the choice of which axis to add to the array. The default is the previous behaviour (axis=0)
• in qp.concatentate, the concatenation of ancillary data is called with axis=0
• added a test to test_factory.py to ensure this was working properly

Code Quality

• My code follows the code style of this project
• I have written unit tests or justified all instances of #pragma: no cover; in the case of a bugfix, a new test that breaks as a result of the bug has been added
• My code contains relevant comments and necessary documentation for future maintainers; the change is reflected in applicable demos/tutorials (with output cleared!) and added/updated docstrings use the NumPy docstring format
• Any breaking changes, described above, are accompanied by backwards compatibility and deprecation warnings

[eacharles]

I added some stuff to deal with the ppf of the spline, which had some numerical issues.