Wave analysis - page 19

 
sak120 писал(а) >>

Marking with structure labels goes instantly without any time consumption (6 rays backwards is enough).

I'm not ready to show the pattern selection algorithm, there are no problems in manual mode. I believe the most appropriate algorithm for patterns is to allow 1 error in the notations, ie exactly one ray to add any notation, so that the full classification was found, ie the answer will be in the form of probability patterns. If one imagines that this does not work, then consider the situation undefined, but this would not be at all frequent.

The partitioning for finding converging triangles or pennants I did is very simple...highlighting breakouts from channels is also not difficult...so why do you have such difficulties with automation...

 
Automating with history checks is a well-established thing...
 
sak120 писал(а) >>

Partitioning with structure labels goes instantly without any time consumption (6 rays backwards is enough).

I'm not ready to show the pattern selection algorithm, there are no problems in manual mode. I believe the most appropriate algorithm for patterns is to allow 1 error in the notations, ie exactly one ray to add any notation, so that the full classification was found, ie the answer will be in the form of probability patterns. If one imagines that this does not work, then consider the situation uncertain, but this would be quite rare.

So the patterns may overlap. The algorithm for identifying zigzag patterns is not a problem.

 
forte928 писал(а) >>
Automating with history checking is a well-established thing.

I'm willing to do wave markup, if it paid for it :), no problem.

A little less information is needed for a trading decision.

 
goldtrader писал(а) >>

Niroba himself put this book on one page:

>> thank you!

 
Integer писал(а) >>

So the patterns can overlap. The algorithm for selecting zigzag patterns is not a problem.

Patterns cannot overlap: the ZigZag is unambiguously and completely broken into patterns. I will now draw a picture in Paint to clarify the situation.

 

The first figure is an ideal situation: wave w(-1) completed the Neely pattern (labelled L3, L5). The next wave w(0) has an impulse characteristic of :5. Wave w(1) seems to complete and starts moving towards w(0). w(0) has started a new upward pattern on Neely. The situation is in buy. The new figure may turn out to be an impulse five or it may be a ZigZag (correction), we don't know.

If we look at the senior frame (senior ZigZag) we can guess.

 
sak120 писал(а) >>

I'm willing to do wave markup, if it paid for it :), no problem.

For a trading solution we need a bit less information.

The information on TF is very noisy, and when using SZ the information decreases by an order of magnitude, so what big volume of data are you talking about, if everything comes down to the fact that the task of constructing just minimize, ie to make the construction of known figures in accordance with the transition between the points....for example we have a figure of a convergent triangle we have 6 points 1-3-5 bottom part of a figure and 2-4-6 top part...it only remains to find out that the figure at the given moment corresponds to the construction with available deviation of the middle point and a convergent triangle...The equation of intersection of lines the problem 8-9 class...

 
sak120 писал(а) >>

The first figure is an ideal situation: wave w(-1) completed the Neely pattern (labelled L3, L5). The next wave w(0) has an impulse characteristic of :5. Wave w(1) seems to complete and starts moving towards w(0). w(0) has started a new upward pattern on Neely. The situation is in buy. The new figure may turn out to be an impulse five or it may be a ZigZag (correction), we don't know.

If we look at the senior frame (senior ZigZag) we can guess.

So we should also move to the senior ZZ for further consideration, what's stopping us?

 
The issue of displaying the higher TF on the lower TF has long been a solved problem, which allows for a data volume of 10,000 points to be displayed on M1 H4//.
Reason: