| tomham |
|
Gold User, Member, TeleChart
|
| Registered User |
|
|
|
|
| Unsure |
|
| Sunday, November 21, 2004 |
| Tuesday, August 6, 2019 11:50:40 AM |
84
[0.03% of all post / 0.01 posts per day] |
|
For those interested, I found the calculation problem with the Read Code above. Here is my solution:
Change the line: Plot = SumSq ^.5 / Period
To: Plot = (SumSq / Period) ^ .5
i.e. take the sum of the squares calculated in the Do Loop in above lines (SumSq), divide by the Period, then take its Square Root.
With this change the Real Code matches perfectly with the calculations in other software.
"Hats Off" to Woden Support for providing the base code for this. StockFinder Rules !
Good Trading to All. .......... Tom ...........
|
I see what you mean Bruce. As soon as I see a ratio (division) and a *100, it's an indication of percent all right. :)
But, the Real Code does not come close to yielding the expected resultant value. When plotted, it is also very erratic, in that there are many spikes, where I see a rather smooth curve of this indicator when plotted else where. Case in point:
The 21 day UI for HISMX -
FasTrack = 2.26
MetaStock = 2.58
This RealCode = 0.49
Checking 3 other funds in two time frames gives similar results. Something is not quite right, but I'll have to dig into it. Thanks for the assistance tho. StockFinder is a very neat program, what a terriffic toolbox. Cheers ............. Tom ..............
|
Hi Bruce (& others) -
I was looking for the Real Code for the Ulcer Index, and found your previous post on it. I did a 'Copy & Paste' into SF5, it compiled and ran fine.
My question is more with the numeric results of the formula, as they don't come very close to the value from another piece of software (that I'm in the process of phasing out of).
This is my understanding of the formula:
Ulcer Index = SquareRoot((The Sum of all R² values) / N)
Where the R is the % amount that the price is below its previous high water mark value, for each day in the series.
Not being to familar with VB, your formula looks pretty representative. The only thing I'm not sure of it that the each R value is a % and not the value of the price difference. Could you take a quick look and verify the logic in its calculation for me ? Thanks.
|
Yes, that's it.
Sounds like I will need to change the logic on it; sort of like removing a "% True" on TC. :)
|
One problem when I created a New Real Code Indicator. After Pasting the code above, and attempting to Save it, I received a "Complie Failed" error message.
In the Error Message window it stated: Name 'Pass' is not declared.
It must be thinking that Pass is a varible, and not a standard Basic command to continue if true.
What should I change to correct this (my Visual Basic skills need help). :) ........ Tom ............
|
|
Thank You Bruce ! This will be a big help in my studies. ........... Tom ..............
|
|
I have written an indicator in TC 2007, and it would be great if I could "port" this formula over to StockFinder 5. (the indicator is my version of the "pocket pivot" from the recent book "Trade Like an O'Neil Disciple").
The output of the formula is "True" (% True in TC terms), so it is binary; "1" if true "0" is false. My Thanks for any assistance. Take Care & Good Trading. .......... Tom ............
(C > C1 AND V > ABS(C1 < C2) * V1 AND V > ABS(C2 < C3) * V2 AND V > ABS(C3 < C4) * V3 AND V > ABS(C4 < C5) * V4 AND V > ABS(C5 < C6) * V5 AND V > ABS(C6 < C7) * V6 AND V > ABS(C7 < C8) * V7 AND V > ABS(C8 < C9) * V8 AND V > ABS(C9 < C10) * V9 AND V > ABS(C10 < C11) * V10) And
((L1<AVGC10.1 or L2<AVGC10.2 or L3<AVGC10.3 or L4<AVGC10.4 or L5<AVGC10.5) Or
(L1<AVGC50.1 or L2<AVGC50.2 or L3<AVGC50.3 or L4<AVGC50.4 or L5<AVGC50.5)) And
C > ((H-L)*.66)+L
|
Thanks for the reply Ken. Reference my issue & your question below:
One issue that I am seeing wit the new SF 5 build update is when I change the content (i.e. rules) of a Condition, that "Shows The True Markers" on the price chart, it no longer actually shows these True Markers (i.e. blank).
It appears that the Watchlist "Buttons" paint O.K., just not the "True Markers".
Now, I may be doing something incorrectly, but I don't see what it could be since the "Show True Markers" option is still set, and only the Condition content was changed.
Any suggestion on what to try, or did anyone else see this with the new build ?
Cheers!
What part of your condition are you changing? ?Is it a RealCode condition?
I have re-started SF5 with the new patch and one of my two combo conditions now paint True on the chart, the other does not, but both do color the buttons in my Watchlist. These are combo conditions, and the changes made to them were deleteing two rules and then adding two. The rules added were not Real Code, but basically a MA cross of TSV and MS (i.e. "canned" indicators).
I've tried re-filering the Watchlist and Refreshing the Chart, no improvement. Perplexing, that one would paint True and the other not, especially when the rules changed are the same in both combo conditions.
I'm happy to try other ideas. Thanks.
|
One issue that I am seeing wit the new SF 5 build update is when I change the content (i.e. rules) of a Condition, that "Shows The True Markers" on the price chart, it no longer actually shows these True Markers (i.e. blank).
It appears that the Watchlist "Buttons" paint O.K., just not the "True Markers".
Now, I may be doing something incorrectly, but I don't see what it could be since the "Show True Markers" option is still set, and only the Condition content was changed.
Any suggestion on what to try, or did anyone else see this with the new build ?
Cheers!
|
|
Super ! Thanks Bruce. it's look'in great now. Take Care.
|
|