Support continuous cell-level characteristics

[arteymix]

•  add a data field similar to what we have for single-cell data vectors
• add a representation
• support INT, BOOLEAN, FLOAT, DOUBLE, etc.

I think it would be a mistake to use a QT for this, but if we need more metadata, I guess we could include a scaleType, etc.

It seems that a boolean would be suitable for encoding masks that would be used for filtering cells. This is relevant to @rachadele.

[arteymix]

For now, we'll be using a categorical CLC for masking cells to include in an aggregate, more details in #1287.

[arteymix]

It might make more sense to have a separate class, or a subclass. I'm not sure what the common interface would be between categorical and continuous CLCs though.

[arteymix]

Got it, we can have a CellLevelMeasurements class that is mirroring our Measurement model, but instead of having a single value, it would have a vector of values, one per cell.