With all due respect, you've requested far more than "be sent in the right direction." You've even lamented the fact that you don't feel you have received straight answers. In fact, when someone 'pointed you in the right direction' you commented on how lengthy the process would be to locate the information yourself. Throughout every Journal I have authored, an atmosphere of collaborative iterative refinement has existed which fostered the sharing of both ideas and coded tool sets. If you want to strike out on your own, fine, but don't hold any expectations that people plan to share their knowledge with someone who appears to be saying, "Once I've learned what I need, I'm taking my toys and going home." People tend to become tight lipped in such an environment. And really, who can blame them? - Spydertrader
with all due respect spyder, did you read my reply to that post? The issue here is not code. It's organizing this mountain of information reporting back to all of you what I find. I'm not interested in hiding any discovery. It's clear I'm in the fact-finding stage right now. Once I am further into the development of auto tapes and channels, my desire is to continue discussion concerning specific, technical aspects of doing that. As far as your journal is concerned, I commend you for building such an impressive volume. But the fact remains that it is unindexed. It is very difficult to find answers in 1000 pages, sifting page by page. It is akin to a doctor trying to look up specific illnesses in a medical encyclopedia that has no index or method of cataloging according to each system of the body. The fact that individuals in this thread come with varying answers to my questions is a testament to the fact that straight, easily accessible answers are hard to come by for this method. My last question is very simple: is a 2-bar pennant sufficient to make a tape? the answer I get is that the formation I am describing is a pennant, specifically FTP and FBP. It's also true that a terrier is a type of dog, but it does nothing to answer my question. I'm not striking out on my own. I'll say it again and again: I am seeking information from the experts to build a model I can demonstrate to the participants in this thread. Code is simply a RESULT of facts and rules organized into algorithms. It is these algorithms I wish to dissect with everyone here. RoughTrader
Yes, that was exactly the thinking I arrived at myself. The reason is intuitive. Tapes always are marked off the RTL. This means ascending lows for up-tapes, and descending highs for down-tapes. an outside bar formation yields ascending highs and descending lows, leaving the formation untapable. The only recourse is to start taping from the outside bar onward. RoughTrader
RoughTrader, Ok, this is good enough for me as I would be more interested in ideas, algorithms, "pseudo-code", etc as opposed to code that I would prolly re-write or an app that I couldn't modify. Although, I do understand the others point of view on this. I've been studying the method for a while and have written some tools specifically geared towards the method. FWIW, I'll throw out some ideas and opinions for you to consider. Regarding channels, I think it's cool you're working on auto-channels but it wouldn't be the approach I would take. I would semi-automate them. Why? 1) While drawing channels can be somewhat mechanical, there are times when experience and discretion come into play. If one draws every mechanical channel, it is my opinion that a certain percentage of them look "ok" but are out of context. Note: I'm all for things being mechanical. 2) Dealing with carry-over (CO) channels from previous day(s) prolly adds more complexity. If you leave out the CO's and don't account for the previous sentiment, your context can be wrong right off the bat. For example, a mechanically drawn dominant traverse is actually a non-dominant retrace. 3) Have you tried to trade mechanically drawn channels at the Forest or traverse level? It is my opinion that one must consider more than channels to base decisions. IMHO, the "interesting" part of this stuff is being able to quantify Change and Continuation (CHG/NOC). Being able to determine CHG/NOC as price approaches the RTL is more important than price breaching the RTL, again, IMO. I believe once you have a good handle on the method, you'll see how important CHG/NOC is, and the need to be able to quantify CHG/NOC on multiple fractals. I would code up a channel "object" that is intially drawn from 3 handrawn points (1, 2, 3). These objects could be intraday or span multiple days. Once instantiated, I would make try to make the object smart enough to handle volatility expansion (prolly easy) and channel "widening". The widening part is intersesting. To do this, one needs to at least look at volume from a Gaussian point of view or a quantifiable method (not pattern recognition) to measure increasing/decreasing dominant/non-dominant volume. I've have some ideas on this that may or may not be worth anything. Anyway, I'm definitely not trying to discourage what you're doing, I'm just throwing a few things out there for you to think about. Regardless, best of luck with your app, spooz
Hi spooz, I'm not sure if I'm doing this correctly, but as far as CO channels are concerned, we're dealing with the closing bar of the last session and the opening bar of the next session, right? My assumption is that the bars are treated as a seamless progression for taping, even if a gap exists. Maybe the attachment gives a good example. (I represent tapes as thin lines traversing from L's to H's and vice versa, and channels as the thicker lines). RoughTrader