Hello Pete,

The following part of your formula could evaluate to ZERO and there is a 
possibility of your formula encountering Divide_By_Zero situations. you may 
want to put safeguard for that.

rDen = (HHV( RSI( 13) , 8 ) - LLV(RSI( 13 ), 8 ));
StochRSI = (rNum/rDen)*100;

You can try:

rDen = (HHV( RSI( 13) , 8 ) - LLV(RSI( 13 ), 8 ));
StochRSI = (rNum/max(rDen,0.00001))*100;

However, my understanding is that Divide_By_Zero will make the formula curios 
till end, whereas you are suggesting that if gets normal after some bar. That 
is beyond my understanding.

With Regards

Sanjiv Bansal

--- In [email protected], "Pete H" <dryhe...@...> wrote:
>
> Drop this formula into a new pane on a daily chart of EWD.
> 
> _SECTION_BEGIN("Stoch RSI");
> rNum = (( RSI( 13) - LLV( RSI( 13 ) , 8 )));
> rDen = (HHV( RSI( 13) , 8 )  - LLV(RSI( 13 ), 8 ));
> StochRSI = (rNum/rDen)*100;
> printf("DTosc: " +MA(StochRSI,13) +"\n");
> 
> Plot(MA(StochRSI,13),"Stoch RSI", colorRed,styleLine);
> 
> Plot(MA(C,13),"MA 13", colorGreen,styleDots);
> 
> Plot(RSI(13),"RSI", colorBlue,styleDashed);
> _SECTION_END();
> 
> Scroll back to 10/29/1996 and notice the plot for the Stoch RSI goes to zero 
> and stays there for the next couple of years. There are a lot of gaps in the 
> data, at lest from my vendor. I believe the gaps are causing the Stoch RSI to 
> drop to zero, but why? This apparently has no affect on the MA of close nor 
> on the regular 13 period RSI. So what's going on and how do I modify the math 
> to give accurate results while preventing the Stoch RSI from dropping to zero 
> due to the missing data?
> Thanks.
> 
>   Pete  :-)
>


Reply via email to