Closed troyraen closed 3 months ago
To be specific, am I correct to understand that that should be done here: https://github.com/mwvgroup/pittgoogle-user/blob/main/templates/bq_elasticc_SuperNNova_schema.json and here https://github.com/mwvgroup/pittgoogle-user/blob/main/classifier/main.py#L120
I ask in the context of setting up the MicroLIA
classifier.
Should diaObjectId
be an integer in the BigQuery table schema?
{
"description": "object identifier or name",
"mode": "REQUIRED",
"name": "diaObjectId",
"type": "STRING"
},
To be specific, am I correct to understand that that should be done here: https://github.com/mwvgroup/pittgoogle-user/blob/main/templates/bq_elasticc_SuperNNova_schema.json and here https://github.com/mwvgroup/pittgoogle-user/blob/main/classifier/main.py#L120
Yes that's correct.
Should
diaObjectId
be an integer in the BigQuery table schema?{ "description": "object identifier or name", "mode": "REQUIRED", "name": "diaObjectId", "type": "STRING" },
Yes, looks like it should be an integer (schema, for reference). Not sure why we have it as a string. It might be that the module is getting it from the message metadata (which are always strings), but even so we should convert back to an integer to match the original schema.
Fixed by #21
It will need to be added in two places: