QuantConnect Community

Thank you for share

Louis,

Nice work. I have been trying for a while to create a breakout strategy a swing trading type algo. I am not the best at programming but I would try to use a universe selection based on solid companies (i.e. growth, revenue fundamentals). With the simple idea of wanting to buy solid companies that have broken out, becuase there is usually a good reason the stock is making a move up as opposed to a stock that meets a specific pattern.

Frank

Hi Frank,

That sounds like a great idea.

Louis

I watched the video to get a better understanding of what you did. I understand how to do the Coarse selection  no big deal,  but how would it flow retrieve the symbols then do all the calculation  in the OnData event? In the OnSecuritiesChanged Event? That is where i get lost

Thanks

Frank

Hi Frank,

I haven't putten too much thought into it, but I'd say the easiest way would be to add some kind of dynamic collection (e.g. list or dictionary) in which you save the calculations for each added symbol. Then you could, for instance, add the new symbol and initial calculations in the OnSecuritiesChanged event, and update the values in EveryMarketOpen.

I hope this helps.

Louis

Hi Frank,

Louis' recommendation above is the standard design pattern to accomplish this. As an example, consider reviewing this strategy in the Strategy Library.

Note that the Ichimoku Clouds strategy warms up the historical data the strategy needs during the SymbolData constructor. Instead of making a History call each day as demonstrated in the algorithm above, we recommend warming up the history and setting up some consolidators.

Best,
Derek Melchin

Derek,Thanks for pointing me in the right direction. I having been using both Quantopian and Quantconnect for a while. For me whenever an algo uses Static Assets (Quantopian term) same as Quantconnect AddEquity I can follow the logic of the algo. I grasp the concept of Quantconnect Coarse and Fine Universe versus Quantopian pipeline and how the OnSecuritiesChanged works. But when it comes the class SelectionData(): and how it works with Coarse Selection to applu your indicators I just can't seem to get it. I have looked at the EMA Universe Boot camp sample but the updating of the indicators is soemthing i struggle with

Hi Louis,

First of all thanks for this example algo and for your ongoing work with your TradeOptions website/youtube. I am a massive fan of the little I have seen so far and will keep learning from everything you share. Really good content.

As Derek Melchin pointed out, I noticed this algorithm performs TWO history requests each day, one for CLOSE and one for HIGH, this could be combined into a single history request in Initialise to "warm-Up" the algo, subsequently accessing the CLOSE or HIGH where needed. 

I believe you could put the ORIGINAL history request in the initialise function to "warm-Up" the algo and then use a rolling window to perform volatility calculations. Do you know how to implement these improvements? 

So far I have been unable to successfully implement a rolling window for volatility calculations...

The problem I encountered was the rolling window returns a QC Indicators object, not an array, so while I could access individual items, I could not access the entire array to perform std calcs - perhaps Derek Melchin can help point us in the right direction? 

I known there is a STD indicator but it would be nice to see how to best perform any calculations on all values within a rolling window of historical data without repeated history requests?

Hi Mark,

First of all, thank you very much for the positive feedback. It is greatly appreciated!

I certainly agree that multiple history calls isn't the most efficient solution but it is an easy solution that even newcomers should understand.

As for calculating volatility, have you tried combining a rolling window and a consolidator? There is an example on the consolidator documentation page. I think this should allow you to access previous values.

I hope this helps.

Hi Frank,

I recommend reviewing the solution file to that bootcamp example. It demonstrates that the indicators are updated by calling their respective Update method.

Mark,

Instead of filling the RollingWindow with indicator objects, fill it with indicator values by using `.Current.Value`. Refer to our documentation for an example.

Best,
Derek Melchin

Derek 

Will do,

thanks

Frank