Closed brohrer closed 9 years ago
A closer look at self.running_activity revealed that it appears to be doing what it was designed for--encouraging exploration of state-action combinations that haven't been tried in a while.
The hub accounts for less than 5% of computation time. Because it's features by actions, rather than features by features, it is far smaller than the matrices in ziptie. Ziptie still accounts for the lion's share of computation, even though it's a lot faster than it used to be.
Profile BECCA and look for parts of the hub that can be sped up using numba. Write the numba functions to accelerated these and test the results.
Also, I don't think running activity in the hub is doing what it's supposed to. I think it's nearly 0 most of the time.