[amibroker] Regularized Fisher Transform (for any argument): Re: The ICycle strange behavior

["DIMITRIS TSOKAKIS" <TSOKAKIS@xxxxxxxxx>]

dalengo,
I certainly agree but this was not my point.
Very few people will read your nice 
http://finance.groups.yahoo.com/group/amibroker/message/62504
for a few days. On the other side, many users will try the 
http://www.amibroker.com/members/traders/05-2004.html
for years and will fail to see a proper ICycle for ^NDX or ^N225.
Thatīs why I believe the members area code should be fixed.
Dimitris Tsokakis  
--- In amibroker@xxxxxxxxxxxxxxx, "dalengo" <dalengo@xxxx> wrote:
> Dimitris and others:
> Please try the function below, it works at any argument
> of "fisher transform"
> -dalengo
> 
> --8<--------------------------------------------------------
> // -- Inverse Fisher transform (regularized) --
> /* Inverse Fisher transform of array x is just y=tanh(x).
> Make the following regularization to avoid exp overflow
> -- dalengo 4/19/04 */
> function InvFisherTfm( array )
> {
> e2y = exp( - 2 * abs(array) );
> tanh = sign(array) * ( 1 - e2y )/( 1 + e2y );
> return tanh;
> }
> --8<--------------------------------------------------------
> // -- Example --
> Plot(InvFisherTfm(MACD()*100),"tanh[100*MACD()]",1,1);
> 
> // -- Check that regularized FishT takes arguments like 1000 w/out 
> a problem
> y0=0;
> for(array=44;array<=1000;array=array+1)
>  {
>  y=InvFisherTfm(array);
>  if(Nz(y)!=0)
>  y0=array;
>  }
>  Title="Maximum Accepted Array = "+WriteVal(y0,1.6);
>  
> 
> --- In amibroker@xxxxxxxxxxxxxxx, "DIMITRIS TSOKAKIS" 
> <TSOKAKIS@xxxx> wrote:
> > --- In amibroker@xxxxxxxxxxxxxxx, "DIMITRIS TSOKAKIS" 
> <TSOKAKIS@xxxx> 
> > wrote:
> > > Tomasz,
> > > After a more detailed examination, the function InvFisherTfm( 
> > array ) 
> > > does not work for array>=45 [the exponential becomes >10^40]
> > 
> > To be more specific, array should be lower than 44.361416 as it 
> comes 
> > from the elementary
> >  function InvFisherTfm( array )
> > {
> >   e2y = exp( 2 * array );
> >   return ( e2y - 1 )/( e2y + 1 );
> > }
> > y0=0;
> > for(array=44;array<45;array=array+0.00001)
> > {
> > y=InvFisherTfm(array);
> > if(Nz(y)!=0)
> > y0=array;
> > }
> > Title="Maximum Accepted Array = "+WriteVal(y0,1.6);
> > 
> > > Of course, many arrays may be higher than 45...
> > > Some initial limitations should be added for regular operation.
> > > Dimitris Tsokakis
> > > --- In amibroker@xxxxxxxxxxxxxxx, "DIMITRIS TSOKAKIS" 
> > <TSOKAKIS@xxxx> 
> > > wrote:
> > > > Tomasz,
> > > > I noticed that the ICycle [from Listing 2 of the recent TASC 
> > > traders 
> > > > tips] has some groups of {empty} values, when applied to an 
> > index. 
> > > > The rest values are +1 or -1 and are not similar to Fig. 4 
red 
> > line.
> > > > Is it normal?
> > > > Dimitris Tsokakis



Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx 
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.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/