Tendential planimetry method - page 9

 

I think looking for worms can be even simpler. We have a nice picture as an XY array. For example, we are at 800 counts. In the picture this datum is marked with a red vertical line. This datum corresponds to some array of "Y" values

If we normalize the difference of near "Y" values to the spread, we will get the following picture for the slice corresponding to 800 values:

x-maximums and we get worms with their "power" characteristics in the end like this picture:Then we simply set one condition to display local maxima and get worms with their "power" characteristics in the end (the name is conventional :o) And so for each X
.

 
vaa20003:
If you could make an indicator that draws worms like in the pictures on page 7 and for different TFs, I think it would be useful.


So, what is the problem? I have previously posted the template for the indicator (with calculation of profile function) in this thread. The profile function used by grasn is a special case of that code. We only need to add the procedure of searching for condensations (i.e. minima of this function) and the indicator is ready. The procedure proposed by grasn can be inserted as well.
 
Mathemat: It is not density or distribution, but a function. Horizontally, the period of the waveform, vertically, its value at a fixed bar.

Approximately so: if this function is monotonous, it is likely to be a trend. If it has a lot of local extrema, it is a flat. Its other parameters can be analyzed correlating them to parameters of similar functions on earlier bars.
Let us extend the thought: if we limit ourselves to a small number of bars but covering a large range of periods (say, a dozen of bars from 3 to 200) and then feed, say, ten rows corresponding to certain bars to the inputs of a neural network that outputs market conditions (+-1 - trend, 0 - flat), then something might work out. And the lag is just right, because it reflects not only the current state of movement, but also the recent past. So lag is not always a bad thing...

P.S. Speaking of worms: information about the position of the harness is not enough. You also need information about which particular waving periods form it. This is the only way to try to determine which TF the harness itself corresponds to.
 

Actually, I found this topic not very interesting from the beginning. Beautiful, but ... I guess the snobbish attitude towards mash-ups has an effect :-) In the end I even lost the thread - I can't understand what you're looking for or what you want to do. And then a thought occurred to me, which I implemented.

Here are two variants of the indicator. One is based on EMAs and the other on SMAs. Each one contains 50 MAs in its flow. The other parameters have been selected so that these two indices correspond to each other completely. That is why the picture is quite suitable for comparing their behavior.

The indicator contains two lines. The red one is trend indicator, value is dimensionless: +1 - uptrend, -1 - downtrend. The blue one - trend strength indicator, has a dimension. For obvious reasons it cannot be normalized. In this picture I have normalized it in a very arbitrary way. Ideally (I think) it would be better to normalize it so it would be in the range (-1,1) for flat and trends that are not of interest for trading. But outside the range - go for it !

It's not that I don't like contramotes, I just haven't met them yet. And I think about the contra-motion as a phenomenon, I think very good. :-)) That said, however, I believe that to expect any kind of predictive power, much less that of price, from a wand, as well as from any other indicator, is a harmful illusion. Everyone agrees that the price can turn in any direction at any moment, so what are we predicting?

IMHO, there is only one more or less valid option. If there is a strong trend, it is able to grind out any counter-trend news and consequently we can count on its inertia. That is the case with the dollar, for example. Its foundation is such, that it can reach 1.6 and 2.0 against the Euro. And those are not factors that can disappear in a day or a month. The Americans will have to inflate a lot to correct their finances, if that is even possible in the foreseeable future. And that trend will not be altered by any news for a long time to come.

So what is needed to be completely happy ? A trend indicator, which newbies always ask about here, and an indicator of its strength, which newbies do not ask about, but which is needed even more, because the stronger the trend, the more likely it is that we will not lose if we go with the trend. Perhaps this option, despite its simplicity, can fulfil these functions.

 
Mathemat:
P.S. Speaking of worms, it is not enough to have information about the position of the bundle. You also need information about the exact periods of the waveform that form it. This is the only way to try to determine which TF the harness itself corresponds to.

If you give the bag weights, it seems that it should be a monotonic function of the period. Then the total weight of the strip will contain information about the period. The band is of course an extensible term, but the profile function can be plotted in density units.

P.S. The actual visualisation of harnesses is of course highly desirable, but the expert can work directly with the density function.

 
What I meant was the following. If you fix a bar and plot "X: waving period -> Y: waving value", the strong resistance/support levels at that curve are the harnesses in the section. Information about which sash periods are part of which harnesses can be important in determining at which particular TF the harness might be significant. Well, let's say a certain average period of wipes included in a harness can correlate with the significant TF on which the harness will work.

A trader, no matter how you look at it, still always trades on some fixed, "trading" TF - even if he uses data from different "analytical" TFs. It took me quite a while to realise this. Likewise worms, which look so chaotic because they can reflect information about significant levels in all TFs at once. By the way, the lag is more for the good than for the bad: it takes some time to identify the trend.
 
Yurixx:

It's not that I have a bad attitude towards contra-motions, I just haven't met them yet. But the contra-motion, as a phenomenon, I think of it very well. :-)) That said, however, I believe that to expect any kind of predictive power whatsoever from a dash, as well as from any other indicator, and even more so than from price, is a harmful illusion. Everyone agrees that price can turn in any direction at any moment, so what are we predicting?

I agree by 200%. The trend should be followed!
But some people (myself included) quite often hammer nails with pliers. Uncomfortable.......... But that's if you don't have a hammer handy and it's the right size.

The indicator (you could have put a code, sorry :)) shows where the price is going and with what "force". But it (the price) can move in any direction at any time.
Worms show at which levels the price can change direction or lag. You will notice that we unconsciously continue the worms to the right "to look into the future".
These are 2 different tools for different purposes.
We play ideally with three tools - entry into the trade, waiting in price movement , exit from the trade.
Here we discuss, in my opinion, the tool "waiting for movement - exit".
And the idea is lost - so a lot of "crazy" ideas are put forward. But all of them are not crazy enough to get the desired result.
 

I haven't read the whole thread, but my first impression. Colour is the 3rd coordinate. You need to build XYZ, and analyze where we are in the ravine, crawling up the ridge or coming down from it. Maybe that will help you. Although I think someone should have suggested that already.

 
Prival:

I haven't read the whole thread, but my first impression. Colour is the 3rd coordinate. You need to build XYZ, and analyze where we are in the ravine, crawling up the ridge or coming down from it. Maybe this will help you. Although I think someone should have suggested it already.

There was a conversation about the machete period. And the colour is the same period painted
 
Mathemat:
What I meant was the following. If you fix a bar and plot "X: waving period -> Y: waving value", the strong resistance/support levels at that curve are the harnesses in the section. Information about what waving periods are included in particular harnesses can be important in determining which particular TF the harness might be significant. Well, let's say a certain average period of wipes included in a harness can correlate with the meaningful TF on which the harness will work.

A trader, no matter how you look at it, is always trading in some fixed, "trading" TF anyway - even if he uses data from different "analytical" TFs. It took me quite a while to realise this. Likewise worms, which look so chaotic because they can reflect information about significant levels in all TFs at once. By the way, the lag is more for the good than for the bad: it takes some time to identify the trend.

In principle I think in a similar way. From my side: flagellation eliminates arbitrariness in the choice of time periods. I'm not so sure about the average period - longer strokes are definitely more important. And vice versa - a jolly crowd of short wagons crossing a "long" bundle can blur the picture. It's on the model that I'd have to spin it. In general, here we can look for if not an invariant, then a temporal fulcrum (you may remember :). It is easy to take into account additional indicators - array MA[i] becomes two-dimensional and you can add as many vectors as you like. The first candidates are periods and weights. As for weights, there is an idea to take the average function of the profile as horizontal (vertical).
Reason: