Titles - generated from user data, not machine readable and not predictable.
Names - generated from metaschema, machine readable and predictable.
We probably want to change behaviour so that tabular data resource pandas DataFrames are populated using names not titles. It does not make sense to use titles as they're unpredictable and thus prevent people from being able to reference by column names in algorithm code, taking away a lot of the advantages of using pandas.
This will have the side effect of making it harder to propagate user-generated titles to schemas in resources that are dependent on input data like fitCurve, fitResiduals. Will have to solve that problem separately after this is implemented.
echus
commented
4 months ago
Titles - generated from user data, not machine readable and not predictable. Names - generated from metaschema, machine readable and predictable.
We probably want to change behaviour so that tabular data resource pandas DataFrames are populated using names not titles. It does not make sense to use titles as they're unpredictable and thus prevent people from being able to reference by column names in algorithm code, taking away a lot of the advantages of using pandas.
This will have the side effect of making it harder to propagate user-generated titles to schemas in resources that are dependent on input data like fitCurve, fitResiduals. Will have to solve that problem separately after this is implemented.