Jack! "Fully automated" is more than some scripts here and there. My take on "fully automated" is that there is software that draws the channel lines and enters/exits at the appropriate points. I guess I shouldn't question amateur semanticists.
you could jerry rig linear regression lines in different time frames. and check which linear regression is holding the best. you could do it on all time frames for every minute of trading hours.
I think you're overcomplicating things Spectre... In the other thread you said something about using different moving averages as well, now you are talking about linear regressions. All you need is price, volume, and channels to use this method. Can linear regression lines clarify anything having to do with the basic concepts at hand in this system? If you are interested in learning the Hershey method I strongly suggest mastering the basic concepts before getting too tool-obsessed They will undoubtedly hinder more than help the learning process. Good luck.
Mak's prv.str.neut.squ spreadsheet. (for IB users. There is also a version for IQFeed.) http://www.elitetrader.com/vb/showthread.php?s=&postid=1046171#post1046171 You have to change the contract dates in cells B3 and B5.
Hi hypostomus, Could you post your eSignal screen layout for using Jack's method? I use QCharts and am switching at the end of month to eSignal for realtime analysis. I already use eSignal for end of day analysis. I like the Tick Loader feature to test myself using Jack's method in simulated realtime. Have you looked at eSignal's API option for developing DOM analysis?
Why should jack pay it forward?. The man has been generous enough to give away this method to anyone who would listen and do the work. Why should he provide the automated code to anyone? So that anyone that is not willing to do the work can reap the rewards? If anything, you could write the 'fully automated' code and give out for free as your way of paying it forward.
I tried out the latest QT patch to see if it had fixed the trend line issues I briefly mentioned in the futures journal. Unfortunately, the problem still exists. I will keep the thread posted. A summary of the issues: - Clone a line and when you try to move the copied line, depending on how close the line is to the original line, the original line gets moved even thought the copied line had the selection anchors - If there is a bunch of lines near each other and one tries to move a line that is selected, usually the selection changes to the line that was created first and moves it instead.
I wrote to them on this same issue. when making multiple trendlines from the same starting point. I must move the original line out of the way, create new line , then move original line back in place. with all of our emails , and the support they have shown, we will most likely see results.