Inference Engine Python API has more simplified and high-level interfaces comparing to C ++ and does not implement C ++ functionality exactly the same, only basic functions required for inference and plugins configurations. That is the reason why you can not find some functions.
The set of functions is going to be extended in the future. What exact features do you miss in the Python API?
Currently batch size can be specified on IR generation stage in Model Optimizer, in python code it's just required to fill input blob properly.
But anyway, analogue of setBatchSize will be added to python API in the future.