10$ for upgrading the indicator - page 2

New comment
 

Hee! You could do a simple and linearly weighted one with a fractional period.

 
Integer писал(а) >>

Hee! You could do a simple and linearly weighted one with a fractional period.

Do you mean adding coefficients to 1? For example, for a period of 3.5 SMA can be written as follows:
a1*Close[3] + a2*Close[2] + a2*Close[1] + a2*Close[0], where a2=1/3.5, a1=1-3/3.5;
I.e. they add up to 1.
Is that what you meant?
 
What else can you suggest? A simple interpolation would be?
------
Piotr, maybe a1*Close[3] + a2*Close[2] + a2*Close[1] + a1*Close[0], where a2=2/7, a1=1.5/7;
Otherwise it turns out asymmetrically ;)
Or at the first index as you suggested, and then further down - with recalculation of coefficients for corner points.
 
Svinozavr писал(а) >>
Integer wrote(a) >>

Hee! You could do a simple and linearly weighted one with fractional periods, though.

Do you mean the addition of coefficients to 1? For example, for a period of 3.5 SMA can be written as follows:
a1*Close[3] + a2*Close[2] + a2*Close[1] + a2*Close[0], where a2=1/3.5, a1=1-3/3.5;
I.e. they add up to 1.
Is that what you meant?


Thought so: (0.5*Close[3] + Close[2] + Close[1] + Close[0])/3.5.

You can also interpolate:

(Close[3]+0.5(Close[4]-Close[3]) + Close[2] + Close[1] + Close[0])/4. In this case it is also possible to specify a fractional offset.
 
Integer >>:


Думал так: (0.5*Close[3] + Close[2] + Close[1] + Close[0])/3.5.
))) Well, that's what I wrote. You get the same coefficients.

You can also interpolate:

(Close[3]+0.5(Close[4]-Close[3]) + Close[2] + Close[1] + Close[0])/4. In this case, you should also be able to specify a fractional offset.
Yeah. But first way is more logical. True, the fractional displacement...
 
Fractional periods can only be spoken of after "analytical continuation" of the inductor formulas into the area of non-integer numbers. This is what should be in the ToR, because such a continuation is ambiguous. If the author cannot explain how, at least let him give an example from another terminal.
 
Mathemat >>:
О дробных периодах можно говорить только после "аналитического продолжения" формул индюкаторов на область нецелых чисел. Вот это и должно быть в ТЗ, т.к. такое продолжение неоднозначно. Если автор не может объяснить как, пусть хоть пример приведет из другого терминала.

Let's remove the ambiguity.
Look at it like a geometry problem...
;)
---shift is known. Squares, too.

 
Suppose the period is a non-integer. What formulas do you propose, avatara:
1. for simple waving
2. for linearly weighted
3. for exponential?
 
Alexei, what's wrong with the calculation I suggested? Do you need an indicator to explain it? )))
 
Mathemat >>:
Допустим, период - нецелое. Какие формулы ты предлагаешь, avatara:
1. для простой машки
2. для линейно взвешенной
3. для экспоненциальной?


Consider that AC for now.
As ordered...
;)
12345678
New comment