Hi , Does anyone know if it is possible to write a scan that covers a few days.

As in the first day is a long white day, the 2nd day is a doji and the 3rd day is a black day

Code i found for this on a forum (((((((CLOSE2>OPEN2*1.0025 and LOW1>CLOSE2)and OPEN1<CLOSE1*1.0012) and OPEN1>CLOSE1*0.9988)and LOW1<CLOSE1) and HIGH1>CLOSE1)and CLOSE*1.0025<OPEN)and CLOSE<CLOSE1)

any help appreciated. I suppose we could write it with the candlestick markers but would that cover a few days doing it that way. I suspect not ?

Jalna that's easy to do using the History functions. Hist() or Ref() or Prev.  Any number of days back to the beginning of chart.

Here's a useful scan called NR4 (Narrowest Range in 4 bars/days).

  H-L<hist(H-L,1) and H-L<hist(H-L,2) and H-L<hist(H-L,3)   

See if you can make one that scans for an NR7 and inside day.

Here's info on how to trade them: http://www.trading-naked.com/insidebarNR4x.htm - http://www.trading-naked.com/insidebarNR4x.htm

Cheers!

but  but but what about my one

I can't see how that is narrow range. : (

Doesn't the above say High minus the low is less than the day before and the high minus the low is less than the 2nd day before and the high minus the low is less than the 3rd day before. So they are getting smaller each day, like a mini wedge

Okay, I have converted, tried and tested your 1. Appears to work and found GCL today April 1st 2008.  C=CLOSE O=OPEN are interchangable in BS.

((((((hist(C,2)>hist(O,2)*1.0025 and hist(L,1)>hist(C,2))and hist(O,1)<hist(C,1)*1.0012) and hist(O,1)>hist(C,1)*0.9988)and hist(L,1)<hist(C,1)) and hist(H,1)>hist(C,1))and C*1.0025<O)and C<hist(C,1));

The narrowest range in 4 is the size of the bar, so if the bar looks smaller than the previous 3 then its doing its job.  You almost had it,  each part is only comparing the 1st bar, not each bar with the other.