v0.10.2

Significant bugs and some small features added in this release.

Bugs fixes:

- With the change from directly accessing raw cursors in pyodbc to using
  sqlalchemy, add_hierarchy() was still attempting to pull a DataFrame
  with a raw cursor and .fetchall(). The fix allows new numbers to be
  added to the local database.

- Although lists of PivotArgs were allowed in the report schema, there
  were instances throughout the codebase that hadn't been updated to
  accept those changes. Multiple pivots per query are now possible.

- Automatic adhoc formatting for pivot tables has been updated to
  include compatibility with multiple columns.

New features:

- Included with the updates for PivotArgs is a new attribute `query`.
  With it, slices to the Output DataFrame can be used instead of the
  original DataFrame.

- When sending a compiled report to a file, unused tabs will be deleted
  prior to saving.

- "rolling_twelve" is a new report configuration keyword replacement.
  This keyword will be replaced by the the year-month 12 months prior to
  the most recent close.