Zero sample correlation does not necessarily mean there is no linear relationship - page 43
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
Because: see picture above.
There are two kinds of pictures - frequency distribution and nuclear mushroom. Why would frequency distribution I nuclear mushroom to calculate QC?
2. Read what Avals writes:
There is a method for estimating the error in calculating the Spearman-Pearson correlation coefficient.
There is a method for estimating the reliability of Spearman and Pearson correlation coefficients.
I am not aware of any mention of the requirement of normality and the impossibility of calculating QC for the original series.
The kind of distribution of the correlation matrix depends on the properties of both series and the relationship between them, i.e. it does not have to be the same for all possible series at all... For SB it is one, for some solar flares another...
The point is that if we take 100 series of satellites of the type I(0) and plot the QC distribution for them, and then integrate these series into I(1) and plot the QC for them, the distributions will be fundamentally different and I(1) will not have the notion of average QC at all, because almost any QC will be average..
If you tell me that the correlation between two price series I(1) is 80% - I will tell you that the correlation between these series is -17% (I gave you the number from scratch). And we will both be right, only I do not even need to count QC, but only to invent any number in the range -1.0 - 1.0, so it makes no sense to talk about QC on I(1) if the probability of any value is equal.
...
I am not aware of any mention of the normality requirement and the impossibility of calculating QC for the original series.
What if there is no distribution at all? What kind of error can there be in this case?
Forget about the distribution - put the values into the formula and calculate the error and reliability of the QC. Why guess with your fingers?
If you use the standard formula, the error is small and decreases proportionally to the root of the row length. C-4 actually did the same thing but through monte carlo. I.e. according to that distribution we can calculate hit interval with any probability (CI), as in those formulas. There is a discrepancy between the formulas and the results obtained by C-4
If you use the standard formula, the error is small and decreases proportionally to the root of the length of the rows. C-4 actually did the same thing but through Monte Carlo. I.e. according to that distribution we can calculate hit interval with any probability (CI), as in those formulas. There is a discrepancy between the formulas and the results obtained by C-4
Once again, what is the argument - QC MAY and MUST be counted from the original rows.
What do you mean by "original rows"?)
For I(1) you can?
So you can for the original ones, but you can't for the non-source ones?
For I(1) can it?
Let's look at it together:
There is my post "CC MAY and MUST be counted by source rows." Now pay attention, question - is there the word ONLY in the meaning of "CC MAY and MUST be counted ONLY on the original rows"?)))