When reading JSON data back into Pandas we have to be careful not to let Pandas infer types or else a column named 'date' with a string '1960' in it will get parsed as a datetime, which then causes it to be persisted incorrectly.
I added a test that initially failed, until I made the correction to how JSON is read in.
When reading JSON data back into Pandas we have to be careful not to let Pandas infer types or else a column named 'date' with a string '1960' in it will get parsed as a datetime, which then causes it to be persisted incorrectly.
I added a test that initially failed, until I made the correction to how JSON is read in.
Fixes #482