Tendential planimetry method - page 6
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
And here are the harnesses on RoundPrice - compare them with Mashka's. Good luck with the trend and big profits.
I find the idea of recursively counting mash-ups on different bars much more interesting.
Regarding display - full picture is of course great, but no matter how you implement it, the computer will load a lot. At the same time, the algorithm for calculating strip boundaries, which I wrote above, does not look too expensive.
For those wishing to understand in more detail what is drawn here is attached a code.
I have made the indicator, while it is drawing the grease density profile for one point. The grossozone is indeterminate units. I remind you that lumps correspond to lows. Possible direction of development in my first (or one of the first) post in this thread. The next challenge is a good algorithm for choosing minima.
For those wishing to understand in more detail what is drawn here is attached a code.
Hi.
Now I think we need to smooth the curve and in the loop count backwards on the number of bars of the prevailing period to get the closest min or max. If we find it - break.
Have a good trend and large profits
Hi.
Now I think we need to smooth the curve and in the loop count backwards on the number of bars of the prevailing period to get the closest min or max. If you find one, break it.
Hi
What is the prevailing period? To find the nearest to the price condensation? But it would be desirable to have all small levels (condensations) - as long as the whole range occupied by the small levels is scanned. Also the procedure of smoothing seems to me somewhat subjective (in terms of choice of parameters). At the moment my thoughts are revolving around zigzag algorithms.
Good luck with the trend and high profits
Likewise :)
P.S. Or does the prevailing period correspond to an absolute minimum in the picture?
to eugenk
Greetings Master Yoda, sent to search for midichlorians of equilibrium, potential energy and memory channels. May the Force be with you.
I don't understand why you limit yourself so much? After all, it's not the channels that are linear, it's your attitude to them that is linear. As for "pissed off", Ostap Bender would recommend treating your nerves with electricity, I dare to recommend some great varieties of terrific ale. But there is a side effect - a hopeless expansion of consciousness.
You're right in stating that the market doesn't sit 70% flat, I've done my research on this and shared a bit here: 'Stochastic Resonance' "grasn 13.10.2007 19:07". That's the whole philosophical sticking point about using the flat model (it's very important to know the statistics of the movement model on which the strategy will be built, highlighted for a reason). But it's completely wrong that linear regression channels cannot be predicted, fundamentally. Unlike worms they can be predicted rather precisely, and there are even phenomenological regularities (I liked the word too much).
As for planimetry - I was also fond of it, but I realized that it is the linear regression channels that solve the problem perfectly, not to mention the simple analytical apparatus. But worms and LR channels have one peculiarity - they do not work independently and it is necessary to find additional criteria, but for worms it is very difficult, unlike LR, I have not managed to do it.
The pictures are really beautiful, really it seems that they are deciphered subconscious waves of average traders, blah blah blah blah and all that:
Well look at those harnesses wide and open, and you'll realise that this harness will wiggle around the midline of the corresponding linear regression channel and taqi, why would you want such a hassle when there's another and there's a great cure for it? What do you want from it - accuracy? You won't get accuracy.
Oh yeah, I completely forgot, what about price movement? And he's on those memory waves ..... (there was a bad word here which I removed before the moderator did :o) :
And there are a lot of such places on the charts (see about philosophy), it is enough to put the "current bar" just before the end of a strong harness (in my terminology - worm) and it is almost guaranteed, without additional reliable criteria to make erroneous conclusions. And all the salt, all the Power is then hidden in these criteria, not in the way the channels are built.
PS: MA is and remains an MA, I do not think that it should be given more than it can give, especially (I repeat), such a problem is perfectly solved by LR channels (wow, I already started repeating myself, that's a wrap).
PPS Besides, MA is very late and the longer the period, the more late and what you call "memory" and see now - it's long gone. Rather oddly, by thinking the market is non-linear (I agree completely) you are essentially attributing to it the presence of long term, strong inertia, which generally speaking does not exist either.
And here is the correctly rotated result of the indicator (the frame is different though)
On the horizontal axis is the width of the price band occupied by the 9 greasers, on the vertical axis is the grease number of the band centre. So on the vertical axis is quite a sophisticated value, it's to confuse everyone :) (and also to at least somehow visualise the primary information in MT).
Now to confuse things a bit - the same data, but vertically the distance in price of the same strip centre from one of the boundaries of the range occupied by the greasers
Levels (condensations of smashes) are seen perfectly, all we have to do is to explain it to the computer :)
Candid, I'm sorry, I must be dumb, but I still don't understand your thoughts on further development. Are you talking about assigning different weights to the wagons and building a quantitative model of the "sluggishness" of the market? Alas, I don't have even the most general ideas of how to take up such a task. ... In addition, there is a strong suspicion that we will run into the same non-stationarity. I.e. observation time, necessary for more or less exact solution of such problem will exceed time of market existence in that state, for which this problem is being solved... Actually, I had a similar idea at the very beginning of learning Forex. Indeed, such things as theories of Elliot, Ghana, Williams, etc. are well known and have been published in huge circulations. Quite a few, if not the vast majority, use them. I.e. they open positions or place pending orders according to their recommendations. This means if we have studied all of these theories, understood their recommendations, and learned their relative prevalence, we can rather accurately forecast where the market crowd order accumulation zones are located. This allows the "masters of puppets" (if they exist) to go on a scavenger hunt of the crowd. For us it shows the same support-resistance levels and the reversal zones. In fact, why does the price accelerate, slow down and reverse? Only because it encounters an increased concentration of supply or demand in certain zones. I used to call it "Lohotronics". Or "Second Order Technical Analysis". Unfortunately, I still don't know how to approach it practically. Although I feel that the theory is correct :) Looks like something similar is on your mind. Am I right ?
Grasn, hello to you too, O Great and Terrible One! Oh Hope and Support of all traders! Oh Thunder and Destruction of all the DTs ! :)))
Alas, I sit without good ale. I haven't got any money yet at work. And to earn like a real man ... Alas and ah. I'm trying, but not very successfully yet. So my consciousness is hopelessly constricted, but all the same, I'm trying to broaden it with long meditations on your pictures :)
So, what I do see as the superiority of worms over linear regression channels.
1) Linear regression channels still need to be constructed. There are countless to construct. And since the number of them is infinite (simply great), the price will surely follow one of them. And a happy trader will reply, "Look, brothers! I told you so", i.e. the matter is not their construction, but the selection criterion. But that is what you are saying. I do not know such criteria. Although it is possible to take a graph. Using the zigzag, or perhaps manually, mark important turning points on it. In their vicinity, set a lot of the linear regression channels and then look where the price moved in reality. Perhaps, having thus reviewed a sufficiently long history, we will be able to obtain some plausible selection criteria. And one should not even demand unambiguousness. For example, from a purely practical point of view I would be quite satisfied with 2-5 channels, even without estimating their probability.
Worms, on the other hand, are built by the simplest and most unambiguous procedure. Arbitrariness is excluded here and there is no headache about criteria of selection.
2) Worms, contrary to linear regression channels, have thickness (not to be confused with the linear regression channel width, these things are different !) and density. This allows (POSSIBLY POSSIBLE, it all still has to be checked and verified) to estimate the probability of a breakdown-bounce. I don't know how to estimate the probability of breakdown of a linear regression channel. Although you may have already done that. For a worm, however, the answer (at least a hint of the answer) lies on the surface.
3) Worms are born and die. For real work at the right-hand edge of the screen, of course, it doesn't matter. But for the study of history it does. And a very big one at that. You can see where and how the worm originated, how it passed, what events it influenced and where it dissipated. The channels of linear regression last for infinity. And not a hint of lifespan or historicity is given. Just one channel, all of a sudden, it is not clear why it is being replaced by another. And the word Suddenly, it seems to me, is repugnant to Nature in general and to the Market in particular. Suddenly, only cats are born :)
Of course there is no ideal anywhere. For example, worms don't work at all on the red (in terms of rainbow coloring in my script) side of the chart. Which is not surprising. The red side is exploration of the new. Worms, on the other hand, rely on previous experience. And the red side has nothing to remember and nothing to refer to. That is their SERIOUSLY flaw. Probably the most serious that I am aware of. Linear regression channels, on the other hand, can be built anywhere. And on the red side of the chart as easy as on the blue. The issue with lagging seems to me to be less clear-cut. Perhaps it is a bad thing, perhaps not. I will have to investigate it as well.
Anyway, glad you paid attention to this topic. Maybe it will help you and us in perfecting your system.
2 eugenk:
You're right in the sense that similar thoughts to yours have been on my mind for quite some time. However, although the top seems to loom ahead, there is still an abyss underfoot. In this theme I was caught by the formal analogy of the picture of dummies with the concept of states that I began to try on the market in the theme "Stochastic Resonance". But here I still have to think and think. So, for the time being I'm just following a less global plan, as described (I had to investigate and clarify it :) in the second post of this thread
lna01 01.11.2007 19:02
You can judge the density of swipes by two indicators: the number of swipes per interval and the interval between swipes. I find the second possibility interesting. That is we calculate a function (interval occupied by n tiles)[(average or median value of those n tiles] over the full range of values (we also order tiles by value), find its minima and draw stripes, say, by half-height. We can take n at a glance about 10. Although in reality both n and the method of defining fringes should be picked up as you go along, looking at the resulting images.
P.S. About non-stationarity it is certainly so, but there is also a memory. I mean I want to think so :)
to eugenk
I'm glad that such a trifle as temporary lack of money hasn't separated you from your sense of humour. But just please don't call me such lofty titles in the future - I'm at least not bad, possibly good. Since reading this, which is about thirty minutes now, I've been sitting red and about to burst with importance, splattering the walls of my house. And if I'm in some way in opposition to the real jedi, then let me be the dark jedi, for some reason I always liked them a lot better in the famous film.
Worms, on the other hand, are built by the simplest and most unambiguous procedure. Arbitrariness is excluded here and there is no headache about selection criteria.
Are you so lazy that it is too much trouble to build a channel? Just kidding, the question comes down to searching for stable channels. And if you are looking for side channels, there is no problem here at all. As for analyzing the history, you have to do the same thing, to find out if you can entrust your capital to these worms. I think I've seen you here: https://www.mql5.com/ru/forum/50458 (Vladislav described his approach well, and I modernized it "a bit")
Rather only the density, because the thickness is determined by this density, and as Mathemat correctly pointed out (and he is the head), it will be difficult to get it right. By the way, there are all sorts of clever image processing methods (channels have to be converted to a matrix, like a picture), perhaps one of them will work for you.
I especially liked this one: "Worms are born and die". Thought about it a lot and cried. It would be interesting to hear a sad ballad about a worm like that. Will you tell me about it sometime at your leisure? And everything in Chaos happens suddenly - it's normal and LR channels also collapse and it's not sudden at all. But the HR channel is still a great forecasting tool (good statistical support), but how are you going to forecast your worms and what are you going to do with them anyway?
Here, I absolutely agree, it's the same with the Sith :o)
The red side is very short MAs?
If the assumption (red, it's very short MAs) is correct, then the red part is what you need. And there's a lot to remember there! The small MAs are built on the same intervals as the long ones (which means they have a sort of memory) - don't make this up, they have a lot to tell you. In the red zone, new channels are born, and they are the ones that change the memory of the old one, change it, shift levels (well, that's just the literary way).
Knowing that the price is currently trading above or below some average mark for some period is probably important (it is this "time-smoothed" information that you see in the form of worms). The question then is, to whom, why, what conclusions does that person draw, and how is he likely to react?
Well, as befits an opposition, I cite without evidence (as you do the merits) a number of shortcomings (this is only the beginning) and bluntly stand on its own (taking into account the experience):
Always glad to hear from you, so, reciprocate. I wanted to organize a separate thread called "market memory", there is something to put, moreover, something to talk about (here you have it regularly, by the way, what is it - "market memory"? :o), but there's a catastrophic lack of time. But never mind...