get_or_create_autopilot_run doesn't support most partitioning settings

[mpkrass7]

Partitioning is set in DataRobot in the following way.

partitioning_method = dr.RandomCV(*args) 

project.analyze_and_model(partitioning_method=partitioning_method, target=target)

An analogous use of get_or_create_autopilot_run would be something like the below:

partitioning_method = dr.RandomCV(*args) 
config = {"partitioning_method": partitioning_method, "target": target)

get_or_create_autopilot_run(analyze_and_model_config=config)

Doing this doesn't work because partitioning method returns an object. The only except we have is that users can pass a datetime_partitioning_config dictionary into the function which sets the config for them.

I propose adding a large mapping file that accepts a dictionary containing arguments for an arbitrary partitioning option with an extra required input to specify what that option is as well as a DataDictionary that lists the potential options for the datetime partitioning config