I've started trying to implement Andrea's Clenow's 'Stocks on the Move' strategy from his book on QC. You can find an implementation of it on Quantopian here (I know there is another official thread but this is the one I followed) - 


I've put where I'm at so far below. One concern I have is QC's ability to handle generating an indicator & filtering on ~700 or so securities (after coarse). The tests alone are currently quite slow but I feel I might not be doing things the most efficient way (I am using self.History query to retrieve historical data per asset to calculate the indicator).

I havn't gone too far into it yet & have only written some skeleton code but I know here are the issues I have for now:
1) Passing the self.History data to the regression_slope function to calculate the indicator

2) Returning the indicator in the dictionary, sorting, then taking the top values for trading & putting them into the self.rebal_list dictionary

3) Updating ATR for the remaining securities in self.rebal_list & sending orders based on the position sizing formula

Any help appreciated, I also figured someone might have already coded this strategy here. I will continue to work on it in the next few days as well.