Are chained ocpu.call() faster than successive ocup.rpc()?

Question

I found opencpu to be significantly slower than the corresponding R function called within R. One of the reasons might be that I exchange data back and forth between the server and the client using:

ocpu.rpc()

every time the user gives an input and pulling back the data on the client when the action is finished.

Could succesive calls of:

ocpu.call() 

lead to a significant speed up?

Answer 1

After quite a lot of testing using both it seems to make no difference. But at least it allows to time the different parts of your request: execution of the function and retrieval of the result more easily.