Numerical series density - page 16

New comment
 
Vyacheslav Kornev:
Okay, I got it. I'll tell you what to do. You're just misleading people. You're not looking for density. You're looking for clusters. When we're working with integers.
Row : 1,2,3,4,5,6 is the density reference. 100%
Row 1,2,_,4,5,6 - is less dense.
A more dense row than 1,2,3,4,5,6. Limited by 6, does not exist.

I'm looking for density of clusters of numbers in a number series - logically correct - I'm not misleading anyone.

In the future will be used non-numbers - so restricted to a numerical series as a benchmark is not decent.


 
I have already explained it to you. The density is different at different points what are you trying to achieve? I have already shown you everything. As far back as 3 pages ago when I explained the bandwidth of a row. You physically CANNOT calculate the density of clusters from EVERY row. Because a number can appear in the series between others and the density will now be different.

Is it not enough for you to use numbers during the day, Max week?

The row must be restricted to 2 sides. If the numbers appear chaotically.

Another issue is only if the numbers are always in ascending/descending order. Then the bandwidth will not change for the numbers on the left.
 
We found all the densities in ascending order when we were looking for the centre of mass


 
Vyacheslav Kornev:
I have already explained it to you. The density is different at different points, what are you trying to achieve? I have already shown you everything. As far back as 3 pages ago when I was explaining row striping. You physically CANNOT calculate the density of clusters from an UNLIMITED series. Because a number can appear in the series between others and the density will now be different.
Is it not enough for you to use numbers during the day, Max week?
The row must be restricted to 2 sides. If the numbers appear chaotically.
Another issue is only if the numbers are always in ascending/descending order. Then the bandwidth won't change for the numbers on the left.

I can see that you have academic knowledge, but having knowledge and applying it are different things.

I admit that my searches have different terminological labels, however, I use labels that are logically justified, even if they conflict with scientific reserved terms - forgive me if this is not convenient, perhaps for you, but I proceed from the fact that one cannot know everything, but one must solve the problem.

Since we have a practical problem, let's discuss it more substantively. Answering the question about the insufficiency of using a single set of numbers - the market changes regularly - the window for deciding one chooses, but I prefer 15 minutes - it means that every 15 minutes I need to search the accumulations and select that most likely will influence the market (this regularity has to be determined, if it exists of course).

What are the characteristics of these clusters in the numerical series:

1. Number of elements

2. Location in the numerical series - it is convenient to limit the limits as percentages

3. Size relative to the whole number series

4. Density - how dense are the numbers in the cluster to each other (different methods of calculation)

Analyzed numeric series will be constantly changing - deltas are not stationary, so the proposed method you seem incomplete - we need a criterion that can filter out some deltas automatically - any ideas?


 
We have 50 cells and 11 cubes with numbers
1, 3, 6, 8, 10, 11, 15, 16, 30, 40,50
V1. The densest clusters are: 10,11 и 15,16.

V2. Less dense are: 1,3 and 6,8,10,11 and 15,16

V3. Even less dense are 1,3,6,8,10,11 and 15,6

V4. Then 1,3,6,8,10,11,15,16.

The bottom line is this. We have picked up the delta. That is, we calculate v2 because in this variant there are the most clusters

Aggregation 1,3 takes 3 cells out of 50, i.e. 1.5 cells per cube.

Cluster 6,8,10,11 takes up 6 cells. And here 1.5 cells per cube. I won't go any further.

You didn't want to make 10 and 11 a separate cluster.




In fact, you should know that we've already calculated the centres of mass in the number series. That is the highest density.

And the numbers around them are within the median delta of all the numbers and form clusters.
 
In short, we calculate the average delta. And see which numbers have the most clusters around them and that's it. The centre of the pile-up is that number
 
-Aleks-:

I can see that you have academic knowledge, but having knowledge and applying it are different things.

I accept that my searches have different terminological labels, however, I use labels that are logically justified, even if they conflict with scientific reserved terms - forgive me if this is not convenient, perhaps for you, but I proceed from the fact that one cannot know everything, but one must solve the problem.

Since we have a practical problem, let's discuss it more substantively. Answering the question about the insufficiency of using a single set of numbers - the market changes regularly - the window for deciding one chooses, but 15 minutes is closer to me - therefore every 15 minutes I need to search the accumulations and select that most likely will influence the market (this regularity must be determined, if it exists of course).

What are the characteristics of these clusters in the numerical series:

1. Number of elements

2. Location in the numerical series - it is convenient to limit the limits as percentages

3. Size relative to the whole number series

4. Density - how dense are the numbers in the cluster to each other (different methods of calculation)

The analyzed numeric series will be constantly changing - deltas are not stationary, so proposed you look incomplete method - we need a criterion that can filter out some deltas automatically - any ideas?


So, please. You will have a limited number of deltas anyway. You can simply set a recalculation period for each bar.
 
Vyacheslav Kornev:
So we calculate the average delta. And see which numbers have the most accumulations around them and that's it. The centre of the cluster is this number.

The average delta, in the previous example, was 122.98 - I thought that was possible, but the figure is clearly significantly different from the selected delta variants.

 
Vyacheslav Kornev:
You're welcome to do so. The range will still be limited. You can simply set a recalculation period for each bar.

Of course the range is limited - each time the limit is different.

However, how to choose the delta range is the question.

 

I've made changes to the script - I've made a more logical calculation of density in the cluster area of the numbers.

Files:
Test_FindOblast3.mq4  25 kb
1...91011121314151617181920212223...26
New comment