Handling Non-Uniform Time Steps During Timetable Schema Modifications

Schema Changes in MATLAB Timetables

MATLAB timetables extend the standard table functionality by associating data with a time vector. While schema modifications—such as adding or removing variables via addvars—are straightforward for standard tables, the behavior becomes complex when the underlying time steps are non-uniform.

When modifying the schema of a timetable that contains irregular sampling intervals, there is uncertainty regarding how the internal indexing and time-alignment are preserved during destructive in-place operations or type casting of columns.

Which mechanisms ensure that the temporal relationship between rows remains intact when redefining the variable types or adding new columns to a non-uniform timetable? Does the timetable subclass introduce specific constraints on schema changes that differ from the base table class?