Epic: upgrade for DESC pipelines

[aimalz]

qp is long overdue for an upgrade that will enable it to serve as a back-end for RAIL, as a continuation of its role in the PZ DC1 paper's metrics. All of qp's open issues still stand, but the most urgent tasks for RAIL are as follows:

• Switch from Python 2 to Python 3 (Code is not currently python3 compatible. aimalz/qp#68)
• Benchmark and improve computational efficiency for large numbers of PDFs
• Make it easy to add new formats with hierarchical classes and a template (Epic: Major refactoring aimalz/qp#94)
• Expand formats to include the native formats of FlexZBoost and MLZ
• Generalize the format conversions (Expand format conversion options aimalz/qp#65)
• Minimize internal conversions when calculating metrics (Better handling of samples aimalz/qp#88)

Since qp is probably transitioning to a DESC code, it also needs to meet higher coding standards:

• Testing (Start unit tests aimalz/qp#3)
• Documentation (Upgrade .travis.yml to make and deploy Note PDF aimalz/qp#62)

The upgrade should be done without inhibiting plans to extend qp's functionality:

• Optimizing quantile spacing (Better quantile spacing aimalz/qp#35)
• Changing the definition of grid evaluation limits (Rethink qp.PDF.limits aimalz/qp#95)
• Adding new metrics (KS test aimalz/qp#47)
• Adding visualizations (Q-Q Plot aimalz/qp#42)

[johannct]

@aimalz @sschmidt23 is this still relevant to organize the work on qp? Some referenced issues are long closed