Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
Quantopian Engineering Team Update - Speedups for Transforms

Hi all,

We have been carefully monitoring the backtester's performance during this alpha release, and we just rolled out the first batch of optimizations based on our findings. Algorithms that access any of our built-in transforms should see significant improvements in runtime, with the greatest benefits accruing to algorithms that access multiple transforms and multiple securities.

So, if you're working on any algorithms that use moving average, volume-weighted average price, or standard deviation, etc., you'll find that backtests run much faster now!

Disclaimer

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by Quantopian. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. No information contained herein should be regarded as a suggestion to engage in or refrain from any investment-related course of action as none of Quantopian nor any of its affiliates is undertaking to provide investment advice, act as an adviser to any plan or entity subject to the Employee Retirement Income Security Act of 1974, as amended, individual retirement account or individual retirement annuity, or give advice in a fiduciary capacity with respect to the materials presented herein. If you are an individual retirement or other investor, contact your financial advisor or other fiduciary unrelated to Quantopian about whether any given investment idea, strategy, product or service described herein may be appropriate for your circumstances. All investments involve risk, including loss of principal. Quantopian makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances.

1 response

Here's an example of an algorithm that's been substantially improved by these changes:
Thomas Wiecki's Dual Moving Averages algorithm for a full 10-year history took
over five hours yesterday. The algorithm accesses two transforms (moving averages of different
lengths) on five different securities. After today's changes, we ran a clone of Thomas' algo on
our staging environment over the same period.

  • Algorithm: Cloned from "Dual Moving Averages - 10 year test not so subpar"
  • From: 2002-01-03
  • To: 2012-07-30
  • Started: 2012-08-23 02:03:37 PM
  • Finished: 2012-08-23 03:15:58 PM

The same test on our upgraded system ran in about an hour and ten minutes,
representing almost a 5x speedup from yesterday.