|
PureBytes Links
Trading Reference Links
|
Terry: from a fast look, you may be right. I did not study the code enough
to see that. Logic comparisons is a good way to zero out some portion of a
calculation depending on circumstances. It might not work here.
Ken
-----Original Message-----
From: spxer [mailto:t.willett@x...]
Sent: Saturday, April 27, 2002 3:50 PM
To: amibroker@xxxxxxxxxxxxxxx
Subject: [amibroker] Re: opt this
Ken, I can't really profess to understand what the original code is
doing, but isn't the Min function going to return a zero in your idea
which will be the new minimum value instead of the actual positive
Minimum value of the original line? Terry
--- In amibroker@xxxx, "Ken Close" <closeks@xxxx> wrote:
> Greg, Terry: I started to try this one but stopped because of other
pressing
> tasks.
>
> My approach was going to be to make a bunch of lines and have a
logic
> comparison in them that would zero when not true and the zero would
mulitply
> out to make the line zero.
>
> So a line like this:
>
> Min(Ref(L,-9)*(pds>=9) + (pds>=9)*Plusval* LinRegSlope(L, pds);
>
> If pds = 8, then the entire line above evaluates to Min( ) * 0 +
> 0*Plusval*LinReg.. all = 0
>
> If pds = 9 or above, then the line would calculate as expected.
>
> Do you think this would work??
>
> Ken
>
>
> -----Original Message-----
> From: greg [mailto:greg.bean@x...]
> Sent: Saturday, April 27, 2002 2:16 PM
> To: amibroker@xxxx
> Subject: Re: [amibroker] Re: opt this
>
>
> Terry,
>
> Sorry, but I think the Plusval suggestion I made is a dud. You seem
to need
> to be able to alter the number of lines according to the pds value.
I'm
> stuck on that one.
>
> Greg
>
> ----- Original Message -----
> From: greg
> To: amibroker@xxxx
> Sent: Saturday, April 27, 2002 2:30 PM
> Subject: Re: [amibroker] Re: opt this
>
>
> Terry,
>
> Just an after thought. Maybe try it first with the green colored
lines in
> the code. You will have to play around with it a bit. Just try a
few things
> and see what happens.
>
> Greg
>
> ----- Original Message -----
> From: greg
> To: amibroker@xxxx
> Sent: Saturday, April 27, 2002 2:23 PM
> Subject: Re: [amibroker] Re: opt this
>
>
> Terry,
>
> I don't know if this will work, but it won't hurt to try.
>
> Plusval=pds-1
>
> There's probably something a lot shorter that will get the job
done. I'm
> new at this . Hmmm, I think I might have something now.
> comes to 6 lines of code if you delete the green lines . They
all seem
> to repeat each other except for minus value after L, .
>
>
> Greg
>
> //pds=optimize("pds,14,1,60,1);
> //pds=14;
> Plusval=pds-1;
> Min(L, Min(Ref(L,-1) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-Plusval) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-3) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-4) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-5) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-6) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-7) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-8) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-9) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-10) +Plusva* LinRegSlope(L, pds),
> Min(Ref(L,-11) +Plusval* LinRegSlope(L, pds),
> Min(Ref(L,-12) +Plusval* LinRegSlope(L, pds),
> Ref(L,-13) +Plusval* LinRegSlope(L, pds))))))))))))));
> ----- Original Message -----
> From: spxer
> To: amibroker@xxxx
> Sent: Saturday, April 27, 2002 1:56 PM
> Subject: [amibroker] Re: opt this
>
>
> Greg, the number of lines of code has to change as well. Ie:
the
> lines where you see the +1 through +13 will change as the
parameter
> changes. A 14 period indicator has 13 lines of code. A 3
peroid
> indicator will have 2 lines of code.
>
>
> --- In amibroker@xxxx, "greg" <greg.bean@xxxx> wrote:
> > Hi, Terry,
> >
> > How about trying this.
> >
> > When you want to optimize. Remove // in front of optimize
line, but
> leave them in front of pds line.
> >
> > When you get your optimized value replace the // in front
of the
> optimize line and remove them from the pds line.
> >
> > Greg
> >
> > +++++++++++++++++++++++++++++++++++++++++++++++
> >
> > //pds=optimize("pds,14,1,60,1);
> > //pds=14;
> > Min(L, Min(Ref(L,-1) +1* LinRegSlope(L, pds),
> > Min(Ref(L,-2) +2* LinRegSlope(L, pds),
> > Min(Ref(L,-3) +3* LinRegSlope(L, pds),
> > Min(Ref(L,-4) +4* LinRegSlope(L, pds),
> > Min(Ref(L,-5) +5* LinRegSlope(L, pds),
> > Min(Ref(L,-6) +6* LinRegSlope(L, pds),
> > Min(Ref(L,-7) +7* LinRegSlope(L, pds),
> > Min(Ref(L,-8) +8* LinRegSlope(L, pds),
> > Min(Ref(L,-9) +9* LinRegSlope(L, pds),
> > Min(Ref(L,-10) +10* LinRegSlope(L, pds),
> > Min(Ref(L,-11) +11* LinRegSlope(L, pds),
> > Min(Ref(L,-12) +12* LinRegSlope(L, pds),
> > Ref(L,-13) +13* LinRegSlope(L, pds))))))))))))));
> > ----- Original Message -----
> > From: spxer
> > To: amibroker@xxxx
> > Sent: Saturday, April 27, 2002 12:40 PM
> > Subject: [amibroker] Re: opt this
> >
> >
> > This is only part of the formula. It is for a 14 period
> indicator. As
> > you can see the lines of code will change for every
parameter
> input.
> > This is known as projection bands by Mel Widner, it is
included
> in
> > metastock. (coded: ProjBandBot(14) which makes it easy to
> optimize)
> ============================= CUT
> =========================================
>
>
> Your use of Yahoo! Groups is subject to the Yahoo! Terms of
Service.
>
>
>
> Your use of Yahoo! Groups is subject to the Yahoo! Terms of
Service.
>
>
> Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|