Here is the link, but I didn't figure it out.
I am running the same query for values I check, and then I have a script transform that looks at those to spit-out the values I want for the different bindings.
Here is a way to check on types in python --- return from a dataset was easy
return [type(columnName[index]]
return [type(hours[1]), type(net[1])]
Or maybe push your data to a table, maybe your data has duplicates or type ordering obstacles?
Or hidden milliseconds when using the datetimes?
These were some gotchas that got me and took a bit anyway.