Open ehneilsen opened 10 months ago
... wouldn't it work with a pd.read_sql command? or are you trying to make sure we're agnostic to datatype of the opsim database, and if so, should we add a command in rubin_scheduler to read an opsim database from disk?
In part I was trying to make it db format agnostic, and in part to make it robust to what SchemaConverter considers the "opsim" format: the way I did it, either could change, and if SchemaConverter were appropriately updated, then the change would propagate here.
Adding a command in rubin_scheduler from disk would, indeed, simplify this code.
Looks fine. I don't understand why you're using SchemaConverter.opsim2obs followed immediately by SchemaConverter.obs2opsim on the same data though. (wouldn't it work with a pd.read_sql command? or are you trying to make sure we're agnostic to datatype of the opsim database, and if so, should we add a command in rubin_scheduler to read an opsim database from disk? )