How to compile data

Discussion in 'Automated Trading' started by game, May 26, 2014.

  1. game

    game

    What do I need to know in order to compile market statistics for any question that I can think of?

    Example: If I want to know the number of 10+ point moves that occurred in a given 2 hour window.

    I am not asking how to analyze the data to discover relationships. I am asking how to automate the compilation of the raw data.

    I am using Sierra Charts. I don't know any programming. I know excel. Where do I begin?
     
  2. 2rosy

    2rosy

  3. Enter the price data sydtematically into a db
     
  4. game

    game

    Thanks
     
  5. Does quandl have intraday data?

    OP. If you are familiar with VBA, you could write loops to process the data statistics. I don't recall Quandl having intraday data, but google has limited intraday data and it's not too easy
    to gather, but you could likely find some excel based approaches to do it with a web search.
     
  6. You could switch to Ninjatrader and it would be like this, they have C# within the product:

    Assumes closes[0] is a 10 minute chart:

    if((Closes[0][0]-Closes[0][12])>10)
    {
    Print("The condition was triggered, close [0] greater " + Time[0]);
    }

    if((Closes[0][12]-Closes[0][0])>10)
    {
    Print("The condition was triggered, close [0] lesser " + Time[0]);
    }
     
  7. game

    game


    Thanks for laying it out. I am going to explore Sierra for a while and see if I can do the same over there. Knowing how to extract data like this is so valuable.
     
  8. tiddlywinks

    tiddlywinks


    Exporting bar-data from SC to CSV for "import" into Excel is a simple matter.
    http://www.sierrachart.com/index.php?l=doc/doc_ImportExport.html

    Alternatively, you can add a "Spreadsheet Study" to your chart and the bar data will populate a SC Spreadsheet. You can then cut and paste into excel.

    Additionally, for a VISUAL ONLY of your example, I've attached a chart of NQ using 2 hours bars. Nothing fancy. A SC default candlestick chart with 2 studies.
    The 2 studies are both "Color Bar Based On Alert Condition"... so 2 instances of that one study. All study parameters are default values. The Alert tab in each study contains an alert formula. The formulas (1 in each of the alert tabs of each study)...
    AND(H-L > 10, C > O) this is the purple bar
    AND(H-L > 10, C < O) this is the yellow bar
    Of course I had to choose the color for each study in the "Subgraphs" tab of each study.
    Alert formula documentation is here...
    http://www.sierrachart.com/index.php?l=doc/doc_AlertCondAndScan.html#AlertConditionFormulaFormat

    Hope that helps
    I really think you're making this too difficult... No formal programming is needed.
     
  9. Learn to program in R or Python...
    After a while you'll look back at excel and wonder why you tortured yourself with it for so long...
     
  10. game

    game

    I have been reading the following book:

    http://www.amazon.com/Data-Smart-Sc...TF8&qid=1401569815&sr=8-1&keywords=data+smart

    While the majority of the book is about developing prototypes in excel, the author concludes by saying that once the basics are covered, its time to learn R. He says that while R is excellent for application it is poor for learning since everything is already loaded up in the libraries and ready to go. Excel meanwhile, allows one to develop a hands on feel for the data. So excel is great for learning and R is great for execution.
     
    #10     May 31, 2014