Fractal theory - page 2

New comment
 
I would like to share a feature I found. You can take a piece on H1, a few waves, go down to m1 and find a similar one there, with the same shape and percentages. It is very accurate, if you find a perfectly similar case. The problem is that the m1 chart is more twitchy. I think it has to do with time sampling. I'll try to discretise it with more than time and see what happens.
 
223231:
I'd like to share my findings. On H1 you can take a slice, a few waves, go down to m1 and find a similar one with the same shape and percentages. It turns out very accurate, if we find a perfectly similar case. The problem is that the m1 chart is more twitchy. I think it has to do with time sampling. I'll try to discretise it with more than time and see what happens.

The feature found is not news to most,

I've observed a couple of times on M1 about 50 bars from D1 (on other periods and even between different pairs there are also many similarities and similarities of figures),

but the trick is not in the observation but in the programming:

1. On what principle and how to add figures to the base.

2. How to learn how to recognize and find figures in the process of formation (how to calculate the percentage of waiting for direction).

3. How to organize statistics for each figure

4. Where to store the database, and how to communicate with it

It only seems simple: "Oh, I have seen the same figures, you can find them, know the future and earn money".

But programming all this is no trivial task. And when it stops drawing what you expect to see is not clear either.

 
ALXIMIKS:

The feature found is not news to most,

I've observed a couple of times on M1 about 50 bars from D1 (on other periods and even between different pairs there are also many similarities and similarities of figures),

but the trick is not in the observation but in the programming:

1. On what principle and how to add figures to the base.

2. How to learn how to recognize and find figures in the process of formation (how to calculate the percentage of waiting for direction).

3. How to organize statistics for each figure

4. Where to store the database, and how to communicate with it

It only seems simple: "Oh, I have seen the same figures, you can find them, know the future and earn money".

But programming all this is no trivial task. And when it stops drawing what you expect to see is also not clear.

Yes, I agree, and before you program it, you must formalize the logic. Just this fact shows the self-similarity of the market. Therefore, I want to move away from the real time and replace it with operational time. I want to see if it gets better. And then I will start programming the search for structural chunks. Basically, it is possible to unify structural chunks at different timeframes and then compare them through correlation.
 
and then tackle the problems further
 

So the subject of fractals has moved into the subject of patterns already ?

 
I can find patterns on a randomly generated graph. It was originally about being able to reconstruct a formula from a fractal pattern, but I don't know how to do it myself and no one advises anything. So I decided to transform the chart to try to calculate this formula with my eyes. It was not about patterns, but about taking the current pattern on a higher timeframe and looking for it on a lower timeframe.
 
Well, basically any piece is a pattern, as long as we look for it as a pattern.
 
223231:
I will share a peculiarity I've found. On H1 you can take a slice, several waves, go down to m1 and find a similar one with the same shape and percentages. It works very accurately, if we find a perfectly similar case. The problem is that the m1 chart is more twitchy. I think it has to do with time sampling. I'll try to discretise it with more than time and see what happens.

Taking a piece and finding a similar piece is nothing more than a pattern search, i.e. a pattern. What does this have to do with fractals? Fractals are based on the operations of addition and multiplication.)

 
elugovoy:

Taking a piece and finding a similar piece is nothing more than a pattern search, i.e. a pattern. What does this have to do with fractals? Fractals are based on addition and multiplication operations, such operations can drive the market to inflation).

but re-read

223231:
I'll share a feature I found. you can take a slice on H1, a few waves, go down to m1 and find a similar one there, with the same shape and percentages. It turns out very accurate, if we find a perfectly similar case. The problem is that the m1 chart is more twitchy. I think it has to do with time sampling. I'll try to discretise it with more than time and see what happens.
 
223231:
I will find patterns on a randomly generated chart. The question was originally about the possibility to reconstruct the formula from the fractal pattern, but I don't know how to do it myself and no one advises me anything. So I decided to transform the chart to try to calculate this formula with my eyes. It was not about patterns, but about taking the current pattern on a higher timeframe and looking for it on a lower timeframe.

Advice can also lead astray.

A fractal is a multiple of some basis, one way or another.

Why H1 and M1, can you explain?

12345678910
New comment