Provide utility function to verify DataFrames

Currently, some tests in feathub verify a pandas DataFrame as follows.

```python
self.assertTrue(expected_df.equals(result_df)
```

Such implementation cannot work in certain cases. For example, if an element in the two DataFrames is a dict whose order of entries is different, they are supposed to pass the test, while the current implementation fails in this case.

Thus we should use a better way to verify DataFrames.

Start by locating the FeatHub tests that compare pandas DataFrames with DataFrame.equals. Review pandas DataFrame comparison behavior for cells containing dictionaries, then add a shared verification utility and update the affected tests so equivalent DataFrames with differently ordered dictionary entries pass.

Written by the indexing model from the issue text.