|
PureBytes Links
Trading Reference Links
|
Very good thinking, Dimitris :-) substituting "1" gives the same result.
Thank you!
What is strange is that I copied the EMPTY definition from Thomasz' sample
C++ code:
#define EMPTY_VAL (-1e10f)
The result of my cut-n-paste programming :-)))
I guess I could have used:
AddToComposite(IIf(IsEmpty(Open),0,1),"_Count","x",7);
But that doesn't work either - so how to spot empty bars?
I searched emails and the AB help, but can't find any clear definition of a
"hole" and the values that must be used to find it. You had quite a
discussion on this with Thomasz some time ago, do you know how holes are
"defined"?
Many thanks Dimitris, you are a great help in keeping the gray matter moving
:-)
Herman.
> -----Original Message-----
> From: dtsokakis [mailto:TSOKAKIS@x...]
> Sent: Sunday, April 28, 2002 11:35 AM
> To: amibroker@xxxxxxxxxxxxxxx
> Subject: [amibroker] Re: Are your Composites accurate???
>
>
> Herman,
> try for this purpose the exploration
> EMPTY = -1^10;
> Filter=Open==empty;
> AddColumn(Open,"");
> Buy=Filter;
> I do not have any in ^NDX.
> But, I run ASE, where a lot of opens==1 and verified my thought.
> Please check your database.You must have some stocks with open==1.
> With your formula, you donīt count these stocks, by assigning a 0 to
> them.
> DT
> --- In amibroker@xxxx, "dtsokakis" <TSOKAKIS@xxxx> wrote:
> > Herman,
> > I just noticed that your EMPTY is -1^10, ie equal to 1.
> > So, your equivalent formula is
> > EMPTY = 1;
> > AddToComposite(IIf(Open == 1,0,1),"~DataPresent","v",3);
> > Do you have in the group of your gif some stocks with open==1 ?
> > This would give some explanation.
> > DT
> > PS The huge negative symbol in AFL is -1e10
> > --- In amibroker@xxxx, "dtsokakis" <TSOKAKIS@xxxx> wrote:
> > > Of course I receive an identical result list with your
> > > EMPTY = -1^10;
> > > AddToComposite(IIf(Open == EMPTY,0,1),"~DataPresent","v",3);
> > > Buy= 0;
> > > f=Foreign("~datapresent","v");
> > > Filter=f!=101;
> > > AddColumn(f,"");
> > > EXACTLY the same results.
> > > To avoid any misuderstanding :your formula works, I just think
> > > Amibroker does not use the open==EMPTY hypothesis, because if the
> > > ADLAC is not present on 16/4/2002, there is no reference for
> ADLAC
> > > this date.
> > > My opinion is from experience, Tomasz knows how AddToComposite()
> > > works.
> > > DT
> > > --- In amibroker@xxxx, "Dimitris Tsokakis" <TSOKAKIS@xxxx> wrote:
> > > > I respectfully disagree. If you are not concerned about bar-by-
> > bar
> > > accuracy
> > > > than you are correct. In that case the "1" method works fine.
> As
> > > long as you
> > > > know that this method will pick up holes of several days but
> that
> > > it will
> > > > not pick up single bar holes.
> > > >
> > > >
> > > > Herman,
> > > > Of course we speak for daily search, bar-by-bar.
> > > > I have in my ^NDX 4 experimental holes on
> > > > 6/1/2000 [1], 15/2/2000 [1] and 1/3/2000[2]
> > > > plus the missing ADLAC after 15/4/2002.
> > > > As you see from the exploration, the population
> > > > is different from 101 exactly these dates.
> > > > I do not understand the conditions of your graph.
> > > > The
> > > > AddToComposite(1,"~count","v");
> > > > Buy=0;
> > > > scans bar-by-bar every stock for each date.
> > > > If the stock is present, it adds an 1 and moves to the next
> stock.
> > > > If ADLAC is not present on 16/4/2002, then the sum will be 100
> > > > for the certain date.
> > > > It is impossible to have a 20% error, there should be another
> > > > reason for your results.
> > > > Dimitris Tsokakis
>
>
>
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
|