Hi Ernie,

Have you looked at my PeakTrough examples at the ZBoard ... I think you will 
find they are a good starting point?

- you need to settle on a definition of a Peak or Trough
- you need to decide which data point, per bar, that you will evaluate
- for accurate reporting of Pk/Tr you need to account for equal highs or lows
- optionally you can evaluate for Major Peaks or Troughs 

I have given examples of all of the above there.


For your own use:

- if you really need lookback versions, and you like my examples as a starting 
template, then you can recode them for your purpose quite easily.

- if you want to quantify Peak or Trough values first you need to define the 
PK/TR and then use Valuewhen(Peak ==1, Close,1)/Valuewhen(Peak ==1, Close,2) 
(or similar) to give you a ROC type answer (I have done this successfully, in 
the past, without the need for loops).

If you are interested, have a go ... if you get stuck post your progress code 
... if I am around I will help or maybe add an example of ROC(Peaks) to the 
Zboard.

In some circumstances you can also simplify the code .. anything that works 
over a reasonable test period is fine.

I call Peaks and Troughs Pivot Points .... note that there are other versions 
of Pivots and Peaks around but I don't have anything to say about them.

http://zboard.wordpress.com/2009/02/20/189/

Brian.


--- In [email protected], "Ernie Newman" <tradi...@...> wrote:
>
> I am trying to design a new Peak Trough Indicator that would 
> detect a peak or trough condition before is has risen to a % value.
> Or one that does not look forward and be able to number the past conditions 
> for use in a scan below is my idea of how this might be programed but the 
> signals are incorrect.    
>  Basically I want to use this as you would a Trough but
> I want to define a low trough value before it makes a 10% advance.
> How can I get the previous Value of this occurance by assigning it a
> number. I am thinking I must use a loop buy not sure?
> I hope I am explaining this correctly.
> THANKS
> Ernie
> 
> 
> // Test Scan to find 10% Lowest Low Troughs by Number
> 
> TrCond = Ref(L,-2) > Ref(L,-1) AND
> Ref(L,-1) < L AND
> Ref(H,-1) < H AND Ref(L,-1) < Ref(HHV(H,10),-2)*.90;
> for( i = 1; i < BarCount; i++ )
> {
> TrCond[ i ] = L[i];
> }
> 
> Filter = TrCond;
> AddColumn(C,"Close",1.2);
> AddColumn(Ref(L,-1),"L-1",1.2);
> AddColumn(TrCond[1],"TR1",1.2);
> AddColumn(TrCond[2],"TR2",1.2);
> AddColumn(TrCond[3],"TR3",1.2);
>


Reply via email to