Discussion of article "Evaluating the ability of Fractal index and Hurst exponent to predict financial time series" - page 4

 


Dear Roman,

let me have 2 notices:

- when interpreting the value of mu (above or below the 0.5 treshold), there must be a mistype.
 In the 'The correlation of the time series nature and the fractal index' section, the relation sign has been exchanged between points number 2 and 3.
 When evaluating the value of H and mu together, the direction of the sign is in the opposite meaning compared to that of the preceding two points.
 Anyway, the Fig. 6 makes it clear that the correct interpretation is that values of mu below (not above) 0.5 correspond to a trend.

- in the attached FRACTAL_upd.zip, there is a corrupted file: fFractalSegmentSeriesAnalysis.mqh. This file ends abruptly in a listing of formal parameters for a method.
 Could you upload an unbroken version of this file?

Thanks,
Paul76

 
Paul7676:


Дорогой Роман,

позвольте мне иметь 2 уведомления:

- при интерпретации значения mu (выше или ниже порога 0,5) должен быть ошибочный тип.
 В разделе «Корреляция природы временного ряда и фрактального индекса» был изменен знак отношения между точками № 2 и 3.
 При оценке значения H и mu вместе направление знака имеет противоположный смысл по сравнению с тем из двух предыдущих пунктов.
 В любом случае, на рис. 6 проясняется, что правильная интерпретация состоит в том, что значения mu ниже (не выше) 0.5 соответствуют тренду.

- в прикрепленном файле FRACTAL_upd.zip есть поврежденный файл: fFractalSegmentSeriesAnalysis.mqh. Этот файл неожиданно заканчивается списком формальных параметров для метода.
 Не могли бы вы загрузить непрерывную версию этого файла?

Спасибо,
Paul76

 
Roman Korotchenko:

Hi Roman, that .mqh file is semi empty too, like the previous.

Anyway, very thanks for the article, you have opened me a new field of vision.

 

Hi

the include file .mqh is corrupted

Just one line :

void fFractalSegmentSeriesAnalysis(double series[], int segmLength, int segmShift

can you provide the correct on ?
 

Nice article, thank you!

 
Roman Korotchenko:

Hi Roman,

That was a great article and we can't wait to try to dig into this by ourselves! However, as already noted the last .mqh file is almost empty.

Can you please upload working code?

 
GREAT article, thank you!!
 

Good afternoon.

Started the indicator, timeframe 1 hour, symbol EURUSD, quotes metaquotes, default settings.

After a few seconds of work, it gives an error array out of range in 'CFractalSeriesSet.mqh' (108,17).

I have this page 108

MuIndexes[ii] = IndexCalculater.Mu;

I have done the sprinting.

16:01:59.441    Fractal Index (EURUSD,H1)       ii = 539
16:01:59.441    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074
16:01:59.441    Fractal Index (EURUSD,H1)       ii = 538
16:01:59.441    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074
16:01:59.441    Fractal Index (EURUSD,H1)       ii = 537
16:01:59.441    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074
16:02:10.682    Fractal Index (EURUSD,H1)       ii = 1610
16:02:10.682    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074
16:02:10.685    Fractal Index (EURUSD,H1)       array out of range in 'CFractalSeriesSet.mqh' (108,17)

I can't understand why the array out of range occurs. The array size is 1074, the index is 1610, where is the overrun?

And it is strange that the indexes are descending, and does not reach zero becomes 1610, before it was all logical.

15:49:38.445    Fractal Index (EURUSD,H1)       ii = 1
15:49:38.445    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074
15:49:38.445    Fractal Index (EURUSD,H1)       ii = 0
15:49:38.445    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074
15:49:47.403    Fractal Index (EURUSD,H1)       ii = 1073
15:49:47.403    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074
15:49:47.403    Fractal Index (EURUSD,H1)       ii = 1072
15:49:47.403    Fractal Index (EURUSD,H1)       MuIndexes.Size() = 1074

Can you tell me what could be the reason for this?

How to fix it?

 
Aleksandr Slavskii #:

I can't understand why the array is overrun. The array size is 1074, index is 1610, where is the overrun?

You answered yourself

 
Rashid Umarov #:

You answered it yourself.

Oops. I see it now.)

For some reason the figure 1610, the brain perceived as 1061 and I am so confused, where is the exit beyond the limits.