Mark Keller
Mark Keller
Let's review this ticket and prioritize it @sfc-gh-hkapre
Hi @TNonet You need to request our optional dependency group when you install `snowflake-connector-python` Instead of `poetry add snowflake-connector-python==2.7.9` use `poetry add "snowflake-connector-python[pandas]==2.7.9"`
I'm sorry @Zaubeerer , I got caught up in some high priority work. I promise you that I'll get to this very soon.
Sorry, the new `write_pandas` functionality has revealed some issues with our internals. So far we know that `datetime`s do not get parsed properly (which is where this error comes from)...
Yes, of course. You can always use the `to_sql` built into `pandas`. It's only the `snowflake.connector.pandas_tools` module that won't work with `datetime`s. I apologize for this issue!
The problem comes from our Parquet parser, so the fastest way to ingest data would be to dump the data into csv files, upload those and then copy those into...
> Hey @sfc-gh-mkeller I found this post on stackoverflow maybe its useful to pinpoint the problem. https://stackoverflow.com/questions/60190210/how-can-i-reliably-use-datetime-values-in-parquet-files-to-fill-snowflake-tabl `fastparquet` seems to solve the issue, so if anyone would like to work...
Hi @Zaubeerer please don't use `-e` it doesn't work with `snowflake-connector-python` for 2 reasons: 1. Using a namespace package. The `snowflake` part of `snowflake.connector` is causing the issue. No one...