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
Let's simplify the problem. 2 one-dimensional arrays of the same length N. The first has speeds, the second has bar numbers. I don't like multidimensional arrays and writing them down. But let's write it as a two-dimensional array for clarity
index 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Values 5,4, 7,8 6,11 4,14 -2,17 3,19 6,24 5,27 9,29 12,32 15,36 13,39 16,43 12,45
You can see that the break occurred on index 9 and from index 10 the average price has changed. How to find /reveal index 9 and 29 bars.
Change detection problem. Perhaps something like CUSUM would work.
Achange detection task. Maybe something like CUSUM would work.
It was the first thing that came to mind)))
I'll probably have to do some checks. If it's more than calculated corridor width by the first sliding window and then sliding window of 3 - 5 values, go to average, nominal and if several windows in sequence, it means corridor, if window parameters in next step go too high, then no corridor.
Let's simplify the problem. 2 one-dimensional arrays of the same length N. The first has speeds, the second has bar numbers. I don't like multidimensional arrays and writing them down. But let's write it as a two-dimensional array for clarity
index 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Values 5,4, 7,8 6,11 4,14 -2,17 3,19 6,24 5,27 9,29 12,32 15,36 13,39 16,43 12,45
You can see that the break occurred on index 9 and from index 10 the average price has changed. How to find/identify index 9 and 29 bar.
I don't know what's in your head and on what principles you consider a channel to be a channel to determine when it ceases to be a channel. You have to try different variants and check visually - whether the result corresponds to the idea of it.
And if simple solutions are satisfactory, then the average direction (i.e. the difference in values on two neighbouring bars) must be counted and if it has changed.Change detection task. Perhaps something like CUSUM would work.
Can you tell me in your own words what CUSUM is and how it's implemented?
===
===
The method is of course a masterpiece, and how deeply thought-out! What the hell - Cambridge! Are you guys all incapable of using your own brains, even formulas in arithmetic only with reference to an authoritative publication?
===
Alexei, and if the raw data can take negative values? That's it, a mathematical disaster?
I don't know what's in your head and on what principles you consider a channel to be a channel to determine when it ceases to be a channel. You have to try different variants and check visually whether the result corresponds to what you imagine it to be.
In everyone's mind, if you draw it you can clearly see where the changes are. On 9. This is in general I agree, it's a problem to solve, but I don't like the CUSUM solution. It then needs filters from single outliers. So far a sliding window. The cumulative calculation is not very good.
In everyone's head, if you draw it you can clearly see where the changes are. On 9. This is generally agreed to be the problem, but I don't like the CUSUM solution. It then needs filters from single outliers. So far a sliding window. Cumulatively calculate is not very good.
So you should not draw, and there must be a dozen or two algorithms in the piggy bank of ideas - apply them and see if the problem is satisfactorily solved.
So you should not draw, but have a dozen or two algorithms in a piggy bank of ideas - apply them and see if the problem is satisfactorily solved.
Well, that's why I'm asking, in case there are solutions). Of course, I will make them myself, but the first solution is not always rational, and I want a nice-looking one. The usual statistics is still for the stationary state, even in the presence of some SB. I just thought that turning a channel with random values has at least some solutions, but I haven't found what I wanted.
So I'm asking if there are any solutions). Of course I will do it myself, but the first solution is not always rational, and I want a nice look. The usual statistics is still for stationary states, even in the presence of some SB. I just thought, that rotation of the channel with random values has at least some solutions, but I haven't found what I wanted.
What do you mean by "solutions"? From Cambridge, by authoritative authorship?
There are 1,000 of these solutions. You just have to sit and try until you see a satisfactory result.
What is the OBJPROP_ANGLE parameter for the trend line ?
For some reason, when a trendline is created, it is always read=0
Although, when rotating the trend line, an angle is displayed, i.e. the developers calculate the angle
What is the OBJPROP_ANGLE parameter for the trend line ?
For some reason, when a trendline is created, it is always read=0
Although, when rotating the trend line, an angle is displayed, i.e. the developers calculate the angle
The angle is in the screen coordinates. The developers have tried, the line holds the angle, but if you change the scale of the graph, the line itself floats away.
It must be assumed that this is a tool for Gann analysis - in fixed coordinates.