|
PureBytes Links
Trading Reference Links
|
Thanks for staying with me, Johan : )
Just to make sure we are "on the same page", let me
use an example. First, I'll re-insert your code here
...
///////////////////////////////////////////////////////////////////////////
Periods = 40;
end = LinearReg( C, periods );
start = LinRegIntercept( C, periods );
slope = LinRegSlope( C, periods );
for( n = x = 0; n < periods; n++ )
x = x + abs( Ref( C, - n ) - ( end - n * slope ) );
Direction = end - start;
Volatility = x;
ER = (Direction / Volatility);
///////////////////////////////////////////////////////////////////////////
If I insert a "static" 40-period linear regression
channel drawing object (with a start date of 16Feb06
and end date of 13Apr06) on the price of symbol SPY,
the end point (the 40th point) on the linear
regression line has a y value of 130.056. If I use the
code Plot(LinearReg(Close, 40),"", colorBlack,
styleLine) on the same chart, the end value for this
"moving" linear regression line (the same as your
"end" variable) is also 130.056. Other values along
the 2 plots will probably not be the same, of course.
It is along the "static" linear regression line that I
want to compute the Efficiency Ratio. At the last/40th
point, the value for scatter = abs(Close - 40th_y) =
abs(Close - end) = abs(128.71 - 130.056) = 1.346. I'm
not using proper AFL here, of course, but i'm trying
to convey my thinking in my own ignorant way. At the
39th point along the 40-period linear regression line,
the value for scatter would be abs(Close - 39th_y). I
want to calculate this value for scatter (the absolute
value of the difference between each day's closing
price and the y value on that static linear regression
line) and then sum the 40 values for scatter to get
the value for Volatility (the denominator in the ER
calculation).
In essence, what I want to do is look at a x-period
segment of various price plots and compare the
volatility-adjusted rate of change of each by
measuring the rate of change along the x-period linear
regression line and the scatter along this line.
Again, i know there are other ways that i might do it,
but this is an approach i wanted to look at, but don't
begin to know how to code.
tia for your (or any one else's) help.
Tim
--- johsun <joh.sun@xxxxxxxx> wrote:
> No, because in this case end will be a different
> value for each of
> the bars in the lookback period; a new endpoint of a
> different
> regression line. You want to calculate the
> difference between the
> previous 40 closes and the current bars 40-bar
> regline. Hope I'm
> making sense;)
>
> Johan
>
>
>
> --- In amibroker@xxxxxxxxxxxxxxx, "timgadd"
> <timgadd@xxx> wrote:
> >
> > Johan,
> >
> > I may have been over-enthusiastic in thinking that
> your code
> worked
> > as I had planned. The numerator (Direction) is
> fine. I'm not sure
> > that the denominator (Volatility) is working as
> planned. It's
> > definitely close.
> >
> > Assuming that the calculation for the "scatter" at
> the last point
> on
> > the linear regression line should be abs(Close -
> End), shouldn't
> the
> > following line return the same value for
> "Volatility" (the sum of
> the
> > scatter) at the last/end point?
> >
> > AddColumn(Sum(abs(Close-end),40),"Volatility
> Check");
> >
> >
>
>
>
>
>
>
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
Please note that this group is for discussion between users only.
To get support from AmiBroker please send an e-mail directly to
SUPPORT {at} amibroker.com
For other support material please check also:
http://www.amibroker.com/support.html
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/amibroker/
<*> To unsubscribe from this group, send an email to:
amibroker-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
|