[BUG] Data dictionary model allows for a continuous BIDS column to have categorical Neurobagel annotations

[surchs]

Is there an existing issue for this?

• I have searched the existing issues

Expected Behavior

"sex": {
        "Annotations": {
            "IsAbout": {
                "TermURL": "nb:Sex",
                "Label": ""
            },
            "Levels": {
                "F": {
                    "TermURL": "snomed:248152002",
                    "Label": ""
                },
                "M": {
                    "TermURL": "snomed:248153007",
                    "Label": ""
                }
            }
        },
        "Description": "There should have been a description here, but there wasn't. :("
    }

is a valid data dictionary entry for a categorical column under our current data model for data dictionaries.
However, notice that this example does not have a "Levels" attribute in the BIDS part (i.e. outside of our "Annotations" key).

Here is what I believe happens:

1. The Column type validates as continuous because there are no required attributes for continuous columns except "Description"and"Annotation"`
   

   bagel-cli/bagel/dictionary_models.py

   Lines 94 to 102 in dd54f14

   	class ContinuousColumn(Column):
   	"""A BIDS column annotation for a continuous column"""
   	units: str = Field(
   	None,
   	description="Measurement units for the values in this column. "
   	"SI units in CMIXF formatting are RECOMMENDED (see Units)",
   	alias="Units",
   	)

   
   (that might be its own issue)
2. The portion inside the "Annotations" key gets validates as CategoricalNeurobagel because it has the levels stuff
   

   bagel-cli/bagel/dictionary_models.py

   Lines 44 to 53 in dd54f14

   	class CategoricalNeurobagel(Neurobagel):
   	"""A Neurobagel annotation for a categorical column"""
   	levels: Dict[str, Identifier] = Field(
   	...,
   	description="For categorical variables: "
   	"An object of values (keys) in the column and the semantic"
   	"term (URI and label) they are unambiguously mapped to.",
   	alias="Levels",
   	)

3. Because we simply say: Annotations have to be either ContinousNeurobagel OR CategoricalNeurobagel, this is all valid:
   

   bagel-cli/bagel/dictionary_models.py

   Lines 77 to 79 in dd54f14

   	annotations: Union[CategoricalNeurobagel, ContinuousNeurobagel] = Field(
   	..., description="Semantic annotations", alias="Annotations"
   	)

I think we have a couple of options here:

• completely abandon the distinction of categorical and/or continuous columns for the BIDS part of the dictionary.
• give the continuous column something that isn't optional (DONE, see [BUG] data dictionary schema let's anything with an annotation pass #151)
• hard-link each column to the specific neurobagel column they should have and only inherit description from the Column class (probably the cleanest)

Current Behavior

No response

Error message

No response

Environment

• OS:
• Python/Node version:

How to reproduce

No response

Anything else?

No response

[github-actions]

We want to keep our issues up to date and active. This issue hasn't seen any activity in the last 30 days.
We have applied the stale-issue label to indicate that this issue should be reviewed again and then either prioritized or closed.

[github-actions]

We want to keep our issues up to date and active. This issue hasn't seen any activity in the last 30 days.
We have applied the stale-issue label to indicate that this issue should be reviewed again and then either prioritized or closed.