I have an Excel based DataSet that I am trying to recreate with a CSV in S3 … I’m manually trying to recreate the dataset meta data but I’m running into issues with dates. I’m trying to use the same date format from my Excel Based dataset (left side) with my CSV in S3 dataset (right side)
If I click Validate it looks to be valid; but when I click Update I get the red error show in the screenshot above.
Can you share an example of a date data point from the dataset? How does the date actually appear in your csv?
I believe this happens because there are no rows to validate against it (i.e they are all null in the preview).
You can manually create a parseDate calculation in the dataset that will take that column and cast it to a date.
Or, if you have it in S3 I would suggest looking to Athena as you can use SQL on your CSVs that can convert this as a date format before it loads into quicksight.
Thank you @Max - I will look at Athena
Though from a product perspective does it make sense to still allow us to still persist the date format configuration … maybe with a warning stating what the problem is.
I definitely agree. I will mark this as a feature request. I believe there is an old question about this in the forum, but will make sure to mark this a feature request again.
