You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
PEtab allows spreading conditions/observables/measurements/... across multiple tables. So far, the different tables of a certain type are merged when creating a Problem. This is convenient for simulation, but pretty inconvenient when loading/modifying/saving the problem, where one usually wants to maintain the old structure.
With the current object model, it should be relatively easy to store the individual tables separately in petab.v2.Problem.
E.g., change Problem.measurement_table from MeasurementTable to list[MeasurementTable] table and introduce a Problem.measurements property that combines them on demand.
Could keep measurement_df from petab.v1 for the combined table, and then a new problem.measurement_tables for the original tables. But then problem.measurement_df should be read-only, unless there is a nice way to edit the original tables via the combined table.
Activity
dilpath commentedon Jul 20, 2025
Could keep
measurement_dffrompetab.v1for the combined table, and then a newproblem.measurement_tablesfor the original tables. But thenproblem.measurement_dfshould be read-only, unless there is a nice way to edit the original tables via the combined table.v2: Don't merge tables when creating `Problem`
v2: Don't merge tables when creating `Problem` (PEtab-dev#405)