Hull Moving Avg does not match results when compared with pine and think script

[serendipity2020]

Hello

I've a pine script in which I'm using below command to get hull moving avg for 55 length.

hma55  = ta.hma(close, 55 )

Now, when I try the same equivalent in ToS and compare the value, it does not match.

def hma55 = HullMovingAvg(close, 55); OR
def hma = MovingAverage(AverageType.HULL, close, 55);

I tried both above but they both give same results, however does not match with result from pine script and miss by a large difference.

Does anyone know why this happens and whats the correct way to convert pine script hma to ToS?

Thanks!

 

[mashume]

@serendipity2020

Neither ThinkOrSwim nor pine script give away the internals of how they calculate the Hull Moving Average. There are many resources on the net which do give the formula for calculation, and you could take that and code it up for a 'third opinion' on which of them is most "accurate". However, given that there is no code available for how Pine Script does it's thing, and without knowing in what way the discrepancy you see is being manifest (you didn't specify what was different, only that they are different) it will be very difficult to impossible to sort out why you see those differences.

-mashume

 

[serendipity2020]

@serendipity2020

Neither ThinkOrSwim nor pine script give away the internals of how they calculate the Hull Moving Average. There are many resources on the net which do give the formula for calculation, and you could take that and code it up for a 'third opinion' on which of them is most "accurate". However, given that there is no code available for how Pine Script does it's thing, and without knowing in what way the discrepancy you see is being manifest (you didn't specify what was different, only that they are different) it will be very difficult to impossible to sort out why you see those differences.

-mashume

Thanks for the reply. Today, I tried it on OXY symbol and the difference was not that big. Pine script was showing 62.56 whereas think script was showing as 61.10

I believe that is acceptable but let me know if its not.

 

[mashume]

So I was curious.

Here's my hand-coded reference HMA from formulas available online:

declare upper;

input length = 55;
input src = close;
plot hma = wma(2 * wma(src, length / 2) - wma(src, length), round(sqrt(length)));

This is then plotted in blue against the ToS built in HMA (which changes color) on the OXY daily chart.

This is interesting, but I have no explanation for it whatsoever.

EDIT
So even when I change the built-in length to match the reference (eyeballing, not actual statistical closeness here) it still doesn't overlap perfectly.

This is reference at 55 and built-in at 52:

-mashume

I once read that a man with one watch always knows what time it is but a man with two is never quite sure.

 

[serendipity2020]

So I was curious.

Here's my hand-coded reference HMA from formulas available online:

declare upper;

input length = 55;
input src = close;
plot hma = wma(2 * wma(src, length / 2) - wma(src, length), round(sqrt(length)));

This is then plotted in blue against the ToS built in HMA (which changes color) on the OXY daily chart.

This is interesting, but I have no explanation for it whatsoever.

-mashume

Wow...interesting...Thanks for the efforts anyways. I believe little bit of offset is ok I guess.