Thanks Bruce...looks like what I wanted
|
I'm looking to build a core holding scan based on
1) A slow, steady uptrend over multiple years
2) A relatively narrow trading range.
Since this is not a trading scan, there is no problem making this multiple scans. However, I need some help fleshing out hte logic.
Let's say I use these definitions:
- multi-year = 2+ year
- slow uptrend = slope in the 10% to 50% range
- relatively narrow = stays within a 10% range over any 1mo period.
How would you approach scanning for these characteristics?
My first thought would be to use the slope of a linear regression channel over a long period of Months. My thought on the range would be that the chart does not violate that channel when set to say 1.0 Standard Deviations.
However, there may be other more elegant solutions.
So, how would I code my approach and is there a preferable approach anyone can offer? (I know Bruce can only provide code for specific scan requests...so the second question is to the community.)
In case it matters, I run 12.6
|
Mine are not resolved...all scans missing now. An hour ago it was the watchlists missing. And when watchlists come back, the contents are sometimes (but not always) missing.
|
Well...they managed to get things working again...for a little while. Now all my watchlists are missing again. So I tried logging out and back in...only to find hte servers are down again (request times out).
Very second rate.
|
Servers down altogether now...can't log in.
|
All my watchlists remain missing...and now all of my scans have disappeared as well. (I had scans when I posted this thread originally.)
Houston...we have a problem.
|
All of my personal watchlists are missing (none visible under Watchlists/Personal) and all watchlists that were open on charts have had their contents replaced by a list of Russell 3000 components.
Please fix whatever problems you caused Friday that had scans not working and then Saturday has made watchlists disappear.
Ver: 12.6.5661.29885
|
I'd like to write a scan (not build off chart) using the StochRSI. Specifically, I want a 7,7,3 StochRSI.
Based on earlier posts (Ver 7 I believe) I developed this code...but it is not working. Can you please help me fix this scan (or start fresh if I'm off-base).
(100 * ((RSI7 - MIN(RSI7,7)) / (MAX(RSI7,7) - MIN(RSI7,7)) + (RSI7.1.1 - MIN(RSI7.1.1,7)) / (MAX(RSI7.1.1,7) - MIN(RSI7.1.1,7)) + (RSI7.1.2 - MIN(RSI7.1.2,7)) / (MAX(RSI7.1.2,7) - MIN(RSI7.1.2,7)) + (RSI7.1.3 - MIN(RSI7.1.3,7)) / (MAX(RSI7.1.3,7) - MIN(RSI7.1.3,7)) ) / 3 ) <=10
This pcf is intended to find StochRSI 7,7,3 <= 10 (but delivers a full range of StochRSI values.
|
Bruce,
I'm trying to replicate an indicator for testing. It's being marketed under the name "X Pattern" and the generic formula is this:
Green line = (indicator Periods - Periods since highest high within Indicator Periods) / Indicator Periods) * 100
Red Line = (indicator Periods - Periods since lowest low within Indicator Periods) / Indicator Periods) * 100
I found the day count thread from way back and want to use that as a base. However, the formula is so complicated I'm not sure I did the conversion correctly. So can you help me nail down a version of the generic formulas above based on "20 indicator periods".
Here is what I have now:
Green Line:
( (20 - ( ABS(H < MAXH20) * (ABS(H1 = MAXH20) + ABS(H1 < MAXH20) * (ABS(H2 = MAXH20) * 2 + ABS(H2 < MAXH20) * (ABS(H3 = MAXH20) * 3 + ABS(H3 < MAXH20) * (ABS(H4 = MAXH20) * 4 + ABS(H4 < MAXH20) * (ABS(H5 = MAXH20) * 5 + ABS(H5 < MAXH20) * (ABS(H6 = MAXH20) * 6 + ABS(H6 < MAXH20) * (ABS(H7 = MAXH20) * 7 + ABS(H7 < MAXH20) * (ABS(H8 = MAXH20) * 8 + ABS(H8 < MAXH20) * (ABS(H9 = MAXH20) * 9 + ABS(H9 < MAXH20) * (ABS(H10 = MAXH20) * 10 + ABS(H10 < MAXH20) * (ABS(H11 = MAXH20) * 11 + ABS(H11 < MAXH20) * (ABS(H12 = MAXH20) * 12 + ABS(H12 < MAXH20) * (ABS(H13 = MAXH20) * 13 + ABS(H13 < MAXH20) * (ABS(H14 = MAXH20) * 14 + ABS(H14 < MAXH20) * (ABS(H15 = MAXH20) * 15 + ABS(H15 < MAXH20) * (ABS(H16 = MAXH20) * 16 + ABS(H16 < MAXH20) * (ABS(H17 = MAXH20) * 17 + ABS(H17 < MAXH20) * (ABS(H18 = MAXH20) * 18 + ABS(H18 < MAXH20) * (ABS(H19 = MAXH20) * 19))))))))))))))))))) ) ) / 20) * 100
Red Line:
( (20 - ( ABS(L > MINL20) * (ABS(L1 = MINL20) + ABS(L1 > MINL20) * (ABS(L2 = MINL20) * 2 + ABS(L2 > MINL20) * (ABS(L3 = MINL20) * 3 + ABS(L3 > MINL20) * (ABS(L4 = MINL20) * 4 + ABS(L4 > MINL20) * (ABS(L5 = MINL20) * 5 + ABS(L5 > MINL20) * (ABS(L6 = MINL20) * 6 + ABS(L6 > MINL20) * (ABS(L7 = MINL20) * 7 + ABS(L7 > MINL20) * (ABS(L8 = MINL20) * 8 + ABS(L8 > MINL20) * (ABS(L9 = MINL20) * 9 + ABS(L9 > MINL20) * (ABS(L10 = MINL20) * 10 + ABS(L10 > MINL20) * (ABS(L11 = MINL20) * 11 + ABS(L11 > MINL20) * (ABS(L12 = MINL20) * 12 + ABS(L12 > MINL20) * (ABS(L13 = MINL20) * 13 + ABS(L13 > MINL20) * (ABS(L14 = MINL20) * 14 + ABS(L14 > MINL20) * (ABS(L15 = MINL20) * 15 + ABS(L15 > MINL20) * (ABS(L16 = MINL20) * 16 + ABS(L16 > MINL20) * (ABS(L17 = MINL20) * 17 + ABS(L17 > MINL20) * (ABS(L18 = MINL20) * 18 + ABS(L18 > MINL20) * (ABS(L19 = MINL20) * 19))))))))))))))))))) ) ) / 20) * 100
Thanks a ton in advance (I know this one is complicated.
|