Kevin N Posted March 30, 2019 Report Share Posted March 30, 2019 Has anyone successfully used MedVedTrader to detect patterns? I know this can be done because I can easily scan for Ascending Triangles over at FinViz. I would like to scan for Ascending Triangles on an intraday chart looking for potential breakouts. Is this beyond the scope of MedVedTraders capabilities? Quote Link to comment Share on other sites More sharing options...
Jerry Medved Posted March 30, 2019 Report Share Posted March 30, 2019 the built in patterns that MT can scan for are candle patterns, not chart patters like this. Triangle detection can be done I am sure though would have to be programmed by you yourself in our advanced mode. We could do it in theory if someone provided specific rules for the detection Quote Link to comment Share on other sites More sharing options...
merlin8121 Posted March 30, 2019 Report Share Posted March 30, 2019 Pattern detection could be a very good addition to the scanner. Ascending triangles, flags etc. 1 Quote Link to comment Share on other sites More sharing options...
Kevin N Posted March 30, 2019 Author Report Share Posted March 30, 2019 That's the problem, I can show you examples (like above) of an ascending triangle, but I can't tell you how to code it. With an ascending triangle, price would hit a price level where it's rejected multiple times, but at the same time it would be putting in higher lows until it gets very narrow between the high and low. At that point, it will usually break to the upside. That would not be hard to code if the pattern was very orderly, but I don't know how to do it because there is no exact amount of time/bars between the first low, next high, next higher low, etc... Quote Link to comment Share on other sites More sharing options...
JimGiles Posted September 10, 2019 Report Share Posted September 10, 2019 For the triangle pattern at least: You can try using a longer time-frame scan than what you are accustomed to and make a rule for the highs and/or lows of the period making progressively lower highs and higher lows than the previous candle period. I understand it isn't as precise as what you are looking for but may be a good workaround to try. Quote Link to comment Share on other sites More sharing options...
Mike Medved Posted September 11, 2019 Report Share Posted September 11, 2019 Yeah, the problem with those patterns is that their definitions are very vague and hard to define in exact terms. What one could do I guess is take the last few (4? 5?) local minima and maxima (which also means you have to define what local minimum/maximum is), run linear regression lines through both sets, and then check that 1. the points on which they are constructed lie close (again, "close" has to be defined) to the lines and 2. they intersect in the future and 3. the slope of the top one is a lot less (1/10th?) than the slope of the bottom one? How does that sound? Quote Link to comment Share on other sites More sharing options...
Kevin N Posted November 8, 2019 Author Report Share Posted November 8, 2019 On 9/11/2019 at 12:59 AM, Mike Medved said: That sounds like it might work, is that something easy to code? I would love to test it out.... Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.