Currently, we have a lot of overlap in the tests in test_data_frame.py and test_pandas_execution.py. This is probably a bad design choice. We should simplify our tests so we have just one set of tests that run with every execution strategy (including SQLite, Pandas, and a mixed strategy).
Currently, we have a lot of overlap in the tests in
test_data_frame.pyandtest_pandas_execution.py. This is probably a bad design choice. We should simplify our tests so we have just one set of tests that run with every execution strategy (including SQLite, Pandas, and a mixed strategy).