Hearst index - page 8
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
These formulas need to be rewritten in discrete form and not get it wrong where.
1. Therefore it is better to have an example in which this Hurst is exactly calculated. So that you can check yourself.
2. I am listening to t and tst and where b appeared Hu=log(R/S)/log(tau/2).
3. I looked on the Net until I found an example.
4. Hurst is calculated using one array. But we need two arrays to calculate the correlation coefficient. That's why we need to decide what to use as the second array.
Only then can we compare
You do not need two arrays to calculate the correlation coefficient between neighbouring samples in the first difference series! If the quotient is denoted by X and the sample length by n, then
Prival, one can use the full formula for finding the correlation coefficient between two BPs, as given in the Encyclopaedia - the result is the same.
There's some nonsense about this shit. I couldn't get 0.5 (although I gave rnd() as input). Unit also failed, although I gave x(i)=i (the series keeps growing)
File attached, Matkad version 14
I was messing around with the PC earlier, so I got 1/2 strictly for the integrated CB. I, however, got an expression for PC myself, but it does not seem to be different from the one you gave me. I will check the formulas later.
Nah, the formulas in the article are glitchy. I got completely wrong results with them (I didn't use your formulas):
It's all bullshit!
It is necessary to get the expressions for the PC yourself. As the basis let's take definition of PC, the essence of which boils down to the fact that the standard deviation of BP at length n grows in proportion n^h, where h=1/2 for inegenerated CB and is not equal 1/2 if NOT random increments are summed up. In other words, if we plot the standard deviation for BP as a function of the TF, we get a straight line with a slope tangent of 1/2 for the random process, <1/2, for the antipersistent and >1/2, for the persistent.
Do you agree with this definition of PC? Don't ask me where I got it from - I reproduced it from memory.
Somehow:
There's some nonsense about this shit. I couldn't get 0.5 (although I gave rnd() as input). Unit also failed, although I gave x(i)=i (the series keeps growing)
File attached, Matkad version 14
You have R = 0 and the Standard Deviation is zero. Accordingly, the ratio R/S=1.
You have R = 0 and the Standard Deviation is zero. Accordingly, the ratio R/S=1.
I'm a little confused as to where this is coming from. I have R=13.5 and the RMS S=2.872 no zeros there.
You do not need two arrays to calculate the correlation coefficient between neighbouring samples in the first difference series! If the quotient is denoted by X and the sample length by n, then
Prival, you can use the full formula for finding the correlation coefficient between two BPs as it is given in the Encyclopedia - the result is the same.
No, that's not it. According to this formula, it turns out that BGS is correlated. All the time, r is around minus 0.5.
Here's the verification code.
Oh, that's cool!
I'll check it out.