|
PureBytes Links
Trading Reference Links
|
From just eyeballing (both eyes) seems that the red line does
approximate a linear regression line of the closes.
None of the other lines comes as close to going through the data points
of the Mov(Typ(),20,S) plot.
Is it the case that this magnitude of discrepancy is due to a number
crunching problem in MS?
Is there that great a difference in the nature of the calculation
between C and MOV(TYP(),20,S)
when TYP=(high + low + close) / 3.
Is the software overwhelmed by these differences in the amount of
calculation?
If yes, Bummer!
Jose wrote:
>Trendchaser, your code is a Linear Regression mixture of the Typ() SMA
>and Close. It doesn't address the original problem.
>
>Ron,
>http://finance.groups.yahoo.com/group/Metastockusers/message/9753
>welcome to one of MetaStock's single precision limitations:
>MS can't handle accurate number crunching in complex processing.
>
>The code below, using and external dll to process the Linear
>Regression values, tracks the Mov(Typ(),20,S) correctly:
>
>---8<--------------------
>
>{ Adaptive.dll (not avail to public) must be in
> ...\MetaStock\External Function DLLs\ folder }
>
>pds:=Input("Linear Regression Trendline periods",2,2520,21);
>
>x:=Mov(Typ(),20,S);
>LRSlope:=ExtFml("Adaptive.LinRegSlope",x,pds);
>LRInd:=LRSlope*Cum(1)
> -LRSlope*ExtFml("Adaptive.Mov",Cum(1),pds)
> +ExtFml("Adaptive.Mov",x,pds);
>
>lastVal:=LastValue(LRInd);
>countback:=LastValue(Cum(1))-Cum(1);
>LR:=
>lastVal-LastValue(LRSlope)*countback;
>restrict:=0-(LastValue(Cum(1))-pds);
>LRT:=Ref(Ref(LR,-restrict),restrict);
>
>LRT
>
>---8<--------------------
>
>Adaptive.dll is not available for public release, so you'll have to
>take my word for it.
>
>
>Now, the same code, using MetaStock's Linear Regression functions,
>plots a Linear Regression line that does not track Mov(Typ(),20,S):
>
>---8<--------------------
>
>pds:=Input("Linear Regression Trendline periods",2,2520,21);
>
>x:=Mov(Typ(),20,S);
>LRSlope:=LinRegSlope(x,pds);
>LRInd:=LinearReg(x,pds);
>
>lastVal:=LastValue(LRInd);
>countback:=LastValue(Cum(1))-Cum(1);
>LR:=
>lastVal-LastValue(LRSlope)*countback;
>restrict:=0-(LastValue(Cum(1))-pds);
>LRT:=Ref(Ref(LR,-restrict),restrict);
>
>LRT
>
>---8<--------------------
>
>
>Linear Regression has been a thorn in MetaStock's side for a while.
>
>
>jose '-)
>
>
>--- In Metastockusers@xxxxxxxxxxxxxxx, "trendchaser2003"
><trendchaser2003@xxxx> wrote:
>
>
>>Try this.
>>
>>Ps:=Input("Mov(Typ() Periods",2,1000,20);
>>Pe:=Input("LinearReg Periods",2,1000,21);
>>x:=Mov(Typ(),Ps,S);
>>x:=LastValue(X);
>>Z:=LastValue(LinearReg(C,Pe));
>>Z:=(Z+X)-Z;
>>Y:=LastValue(Cum(1))-Cum(1);
>>U:=Z-(LastValue(LinRegSlope(C,Pe))*Y);
>>R:=0-(LastValue(Cum(1))-Pe);
>>A:=U+Ref(C,R)-Ref(C,R);
>>Range:=LastValue(Mov(H,Pe,E)-LastValue(Mov(L,Pe,E)));
>>a+(16*(0.125*Range));
>>a+(8*(0.125*Range));
>>A;
>>a-(8*(0.125*Range));
>>a-(16*(0.125*Range));
>>
>>Trendchaser
>>http://www.freewebs.com/trendchaser/
>>
>>
>
>
>
>>>--- In Metastockusers@xxxxxxxxxxxxxxx, karile <karile@xxxx>
>>>
>>>
>wrote:
>
>
>
>>>Hi,
>>>
>>>
>
>
>
>>>I have a code which calculates a linear regression (see
>>>
>>>
>below).
>
>
>
>>>What I would like to have is a linear regression of a 20 SMA
>>>( (mov(typical(),20,s) ) instead of the close.
>>>Could anyone adjust the code below to perform that ?
>>>
>>>
>
>
>
>>>Here is the code :
>>>
>>>
>
>
>
>>>Pe:=Input("Periods",2,1000,6);
>>>Z:=LastValue(LinearReg(C,Pe));
>>>Y:=LastValue(Cum(1))-Cum(1);
>>>U:=Z-(LastValue(LinRegSlope(C,Pe))*Y);
>>>R:=0-(LastValue(Cum(1))-Pe);
>>>U+Ref(C,R)-Ref(C,R);
>>>
>>>
>
>
>
>>>Thanks for your help,
>>>Regards,
>>>
>>>
>
>
>
>>>Karile
>>>
>>>
>
>
>
>
>
>
>Yahoo! Groups Links
>
>
>
>
>
>
>
>
------------------------ Yahoo! Groups Sponsor ---------------------~-->
Buy Ink Cartridges or Refill Kits for your HP, Epson, Canon or Lexmark
Printer at MyInks.com. Free s/h on orders $50 or more to the US & Canada.
http://www.c1tracking.com/l.asp?cid=5511
http://us.click.yahoo.com/mOAaAA/3exGAA/qnsNAA/zMEolB/TM
---------------------------------------------------------------------~->
Yahoo! Groups Links
<*> To visit your group on the web, go to:
http://groups.yahoo.com/group/Metastockusers/
<*> To unsubscribe from this group, send an email to:
Metastockusers-unsubscribe@xxxxxxxxxxxxxxx
<*> Your use of Yahoo! Groups is subject to:
http://docs.yahoo.com/info/terms/
Attachment:
Description: ""
|