Libraries: Statistical Functions - page 2

New comment
 

Well, I must admit that I'm not sure why it's different, but as I can obtain similar result in excel,
presented article cannon be wrong. It's rather some difference in calculating t value and in the set of those critical valuse.
If you would use t value from matlab or mathematica and compare it to critical values presented in the article,
the test result would be the same as for t value calculated with the method presented in the article.
However, considering that all those critical values are negative I would rather believe that implemented method is more suitable for them
(result obtained in mathematica is too big in comparison).

Of course I will try to look up for another methods of calculating t value to find out why those results are so different.

Regards,
Herajika 

 
Haruna Nakamura:

Well, I must admit that I'm not sure why it's different, but as I can obtain similar result in excel,
presented article cannon be wrong. It's rather some difference in calculating t value and in the set of those critical valuse.
If you would use t value from matlab or mathematica and compare it to critical values presented in the article,
the test result would be the same as for t value calculated with the method presented in the article.
However, considering that all those critical values are negative I would rather believe that implemented method is more suitable for them
(result obtained in mathematica is too big in comparison).

Of course I will try to look up for another methods of calculating t value to find out why those results are so different.

Regards,
Herajika 

Hello Haruna Nakamura,

I'm wondering if there is a small error at this line of code, used for calculating the t-value:

tValue=corrCoeff*MathSqrt((n-2)/1-MathPow(corrCoeff,2));

instead of

tValue=corrCoeff*MathSqrt((n-2)/(1-MathPow(corrCoeff,2)));

Regards,
Malacarne

 
Yes, you're rigth. Thank you for noticing it!
I uploaded edited version and it should appear when accepted by moderator.
Thank you once again!
 

Hello Guys,

 

Am new to all of this! Can you someone advise how can I implement into MT5 and use this script? Thanks 

 

First of all it's a  head file. You cannot #import template. So It's include-file.

Then function "T mean(T &arr[])" must return double in any case of T-templates. Mean of 1 and 2 is 1.5 but not 1.

 
Firstly, this is an include file, secondly, the first function T mean(T &arr[]) is not correct, because it should return double, not T. The average of an array of integer doubles.
 
Orangetree:
Firstly, this is an include file, and secondly, the first function T mean(T &arr[]) is incorrect because it should return double, not T. The average of an array of integer doubles.
Do you think Haruna Nakamura knows Russian?
 
Evgeny Belyaev:
Do you think Haruna Nakamura knows Russian?
I wrote to him in English
 

Could someone please make an example with engleGrangerTest?

I tried without results.

 

Hello @Haruna Nakamura,

It seems that the STD function is for population and not for sample. Using this function inadvertently may lead to erroneous conclusions. I would suggest changing the function name and / or adding the function to the sample standard deviation.
By the way,  the MQL5 already have in it's Math library, the function MathStandardDeviation().

123
New comment