This message was deleted.

There's shouldn't be a limit of 50 chars, at least not one I'm aware of. Where are you seeing this?

hm, I am trying to groupby by columns that have names with dots but dont represent any table names (eg some.column.name) which confuses somehow the daft's sql parser, so I tried to escape them with ' and noticed that columns with names longer than 50 chars get cut off at 50th character, weird

just to confirm, is this only an issue with daft sql? are you able to perform the operations using the dataframe api without issue?

I am using dataframe api not sql api

are you able to create a MRE (minimum reproducible example)?

I am porting some data mapping sql queries from sqlachemy to daft, the resulting dataframe looks like this

df = daft.from_pylist(
        [
            {
                "row_id": 13,
                "subject_id": 16,
                "Encounter.class.code": "IMP",
                "Encounter.diagnosis.2.condition.identifier.value": "103251_admission",
                "Encounter.class.system": "<http://terminology.hl7.org/CodeSystem/v3-ActCode>",
                "Encounter.hospitalization.dischargeDisposition.coding.0.code": "home",
                "Encounter.subject.identifier.value": 16,
                "hadm_id": 103251,
                "Encounter.identifier.0.value": 103251,
                "Encounter.length.value": 188160.0,
                "Encounter.class.display": "inpatient encounter",
                "Encounter.hospitalization.admitSource.coding.0.display": "Medical Practitioner/physician referral",
                "Encounter.hospitalization.admitSource.coding.0.code": "mp",
                "Encounter.hospitalization.dischargeDisposition.coding.0.display": "Home",
                "Encounter.period.start": datetime.datetime(2178, 2, 3, 6, 35),
                "Encounter.diagnosis.1.condition.identifier.value": "70_103251_3",
                "Encounter.period.end": datetime.datetime(2178, 2, 5, 10, 51),
                "Encounter.diagnosis.0.condition.identifier.value": "9679_1_103251",
                "Encounter.hospitalization.admitSource.text": None,
            }
        ]
    )

groupby_cols = ["subject_id", "hadm_id", "row_id"]
agg_cols = [col for col in df.column_names if col not in groupby_cols]
aggs = [daft.col(col).list_agg().alias(col) for col in agg_cols]

hmm that example runs just fine for me.

╭────────────┬─────────┬────────┬─────────────┬────────────────────────┬───────────────────────────────╮
│ subject_id ┆ hadm_id ┆ row_id ┆      …      ┆ Encounter.period.start ┆ Encounter.subject.identifier. │
│ ---        ┆ ---     ┆ ---    ┆             ┆ ---                    ┆ value                         │
│ Int64      ┆ Int64   ┆ Int64  ┆ (14 hidden) ┆ List[Timestamp[us]]    ┆ ---                           │
│            ┆         ┆        ┆             ┆                        ┆ List[Int64]                   │
╞════════════╪═════════╪════════╪═════════════╪════════════════════════╪═══════════════════════════════╡
│ 16         ┆ 103251  ┆ 13     ┆ …           ┆ [2178-02-03 06:35:00]  ┆ [16]                          │
╰────────────┴─────────┴────────┴─────────────┴────────────────────────┴───────────────────────────────╯

(Showing first 1 of 1 rows)