Override API

Synth builds statistical models of data automagically from scanning real data. That great strength is sometimes also a weakness and it is important to be able to manually specify additional constraints and custom high-level logic on top of the statistically-inferred information.

The Override API is the subset of API functionalities of synthd that allows for overriding statistically-inferred data models. Effectively, it lets you specify the bounding box in which the core ML-driven data model can freely train. This includes specifying one-to-one, one-to-many, many-to-one relations between collections; date time formats; custom string generators; etc.

For more on how synthd handles overrides, see Overrides.

OverrideClient(client)	Base class for the Namespace API.
OverrideClient.put_override(*args, **kwargs)	Override a schema node.
OverrideClient.delete_override(*args, **kwargs)	Delete a schema node.
OverrideClient.set_optional(*args, **kwargs)	Make a field node optional or required.
OverrideClient.optionalise(*args, **kwargs)	Make a field node optional (not required)
OverrideClient.unoptionalise(*args, **kwargs)	Make a field node required (not optional)

synthpy.client.namespace.NamespaceClient.rollback_namespace synthpy.client.override.OverrideClient