Join GitHub today
GitHub is home to over 40 million developers working together to host and review code, manage projects, and build software together.Sign up
Time Range filter gets Tz-aware datetime.datetime cannot be converted to datetime64 unless utc=True #8943
A clear and concise description of what the bug is.
In PostgreSQL, my products are associated with an industry. I wanted an area chart to see how each industry is trending over time, but I get an error WHEN the time range is greater than a month
I should be able to crate an area chart without filtering.
Get an error
2020-01-09 11:03:17,062:INFO:root:SELECT industry AS industry, DATE_TRUNC('week', creation_time) AS __timestamp, COUNT(*) AS count
How to reproduce the bug
Create a table in postgresql
CREATE TABLE public.test_tz (
Populate the table somehow, at one point the error gets triggered. In my case it's filtering by week above a month.
Ubuntu under Windows (WSL)
Make sure these boxes are checked before submitting your issue - thank you!
I have found a similar issue stating that's it's fixed. It's partially fixed (I guess)
It's very weird that filtering fixes it. I guess some of the entries are on a date. What's ever more weird is that when you state timestamptz in Postgresql it automatically normalzed all entries as UTC (which is odd since I wanted to keep the timezone information, but expected). So Panda insisting on specifying an utc parameter is very odd indeed.