[amibroker] Re: Walk Forward Issue

2009-07-15 Thread brianw468
ntical results. If not, why not? > > The secret to debugging is "divide and conquer", just "one step at a time". > -- Keith > > brianw468 wrote: > > > > > > Oops, > > This message became a mess as Yahoo apparently removed most of the &g

[amibroker] Re: Walk Forward Issue

2009-07-13 Thread brianw468
to get a normally distributed sample. Your next OOS period would then be less > likely to be based on an IS fluke. > > Mike > > --- In amibroker@yahoogroups.com, "brianw468" wrote: > > > > Hi > > I recently tested a trading system but got some strange r

[amibroker] Re: Walk Forward Issue

2009-07-13 Thread brianw468
amibroker@yahoogroups.com, "brianw468" wrote: > > Hi > I recently tested a trading system but got some strange results from the WF > test. The WF process used exhaustive optimization. > An extract from the results (too large to reproduce in full) is:- > ItemProfit

[amibroker] Walk Forward Issue

2009-07-13 Thread brianw468
Hi I recently tested a trading system but got some strange results from the WF test. The WF process used exhaustive optimization. An extract from the results (too large to reproduce in full) is:- ItemProfitNo. Trades IS1 533 1 OOS11126 2 IS2 474 1 OOS

[amibroker] Re: ZigZag code

2009-06-26 Thread brianw468
Hi Here is the updated code as promised. It corrects several things that I believe were errors in the original. All comments welcomed. Brian PS It hasn't copied across too well from Amibroker, so editing line returns might be required! // Zig Zag Code P = Param("Percent",3,1,5,0.01);

[amibroker] Re: ZigZag code

2009-06-23 Thread brianw468
Hi I am interested in this function also, but on inspection found a few things that I believe are errors. When I applied it to a test data set it missed a few valid highs and lows. Looking at the code I saw that when testing for a swing High it (correctly) tests whether this is > End_Val_HL*Wav

[amibroker] AFL Editing

2009-06-21 Thread brianw468
Hi, I'm trying to edit an existing AFL program, but find that when I attempt to insert text within a line, the new text replaces the existing text, rather than pushing it sideways. It should be simple to change this, but I can't find the method. Any suggestions please? Brian

[amibroker] Re: HELP: Code Compaction

2009-05-11 Thread brianw468
Is this more compact again? > _SECTION_BEGIN("code compaction"); > Len = Param("length",21,1,200,1); > V1 = MA(C,len)*1618; V2 = Ref(V1,-5); dn = IIf(V1V2,V1-V2,0); > Plot(dn, "", colorRed, styleThick); > Plot(up, "", colorLime, styleThick); > _SECTION_END(); Brian --- In amibroker@yahoogroup

[amibroker] "Simple" Backtesting

2009-05-09 Thread brianw468
I hope this isn't a dumb question , but I've had some very strange results with what should be a really simple test (as per Howard Bandy's example on p78 et seq in Intro to AB, but with different Buy & Sell signal generation). Simple set up - Long positions only. No commissions, no stops, Buy and

[amibroker] Re: Simple Question

2009-05-01 Thread brianw468
Status("lastvisiblebar") returns -1. Presumably this is because the code is > not being run in the context of the chart. However, if you simply save the > changes, your chart will correctly update to reflect your code. > > Mike > > --- In amibroker@yahoogroups.com,

[amibroker] Simple Question

2009-05-01 Thread brianw468
Can someone please advise where the error is in the following simple code:- fbr = Status("firstvisiblebar"); Lbr = Status("lastvisiblebar")-1; x0 = fbr; x1 = Lbr; y0 = H[fbr]; y1 = L[Lbr]; Line = LineArray(x0,y0,x1,y1); Plot(Line,"Li

[amibroker] Re: ROUND function - How to RoundUp ?

2009-04-21 Thread brianw468
Try round(sqrt(CountAll)+0.5)). --- In amibroker@yahoogroups.com, "gmorlosky" wrote: > > What is a an easy way to RoundUp, instead of just Round ? > I take the square root of a number (say 85) and then it's square root (9.2) > and then round (9) to create a square matrix (total of 81), which c

[amibroker] Simple Problem with "flip" function and/or "Plotshapes"

2009-01-26 Thread brianw468
The following simple program illustrates an intriguing problem I encountered when trying to develop a complicated trading system. //Test Program fast = MA(C,15); slow = MA(C,50); Buy = Cross(fast,slow); Sell = Cross(slow,fast); Buy = Flip(Buy,Sell);

[amibroker] Re: Optimization Question

2009-01-08 Thread brianw468
r be in Buy territory or be of No ( 0 ) Use. > > Another possibly more sophisticated way to attack this might be to have a voting mechanism or a weighted voting mechanism ... > > - Original Message - > From: brianw468 > Date: Thursday, January 8, 2009 4:03 pm > Subje

[amibroker] Re: Optimization Question

2009-01-08 Thread brianw468
Employing "Use" variables as shown below is a clever approach to sorting out useful vs non-useful indicators, BUT am I missing something basic here? The "OR" parts of each condition seem wrong. Surely the code should read (e.g):- RSIx>Ref(RSIx,-1) AND RSIUse == 1; and similar for the other indic

[amibroker] AFL coding

2008-12-02 Thread brianw468
Can anyone help with the following, please:- 1. I am developing an AFL to generate buy and sell signals by combining different indicators with variable parameters (to be optimised). I would like to generate a plot with price and one indicator in the top pane, and the other indicators in separate

[amibroker] Re: Recursive Boolean Expressions... Possible?

2008-09-19 Thread brianw468
; > This is equivalent to the loop code: > > varx[i] = varx[i-1] + 1; > > Regards, > GP > > > --- In amibroker@yahoogroups.com, "brianw468" wrote: > > > > Hi, > > Can you not solve the problem by (effectively) re-defining varx > > wit

[amibroker] Re: Recursive Boolean Expressions... Possible?

2008-09-18 Thread brianw468
Hi, Can you not solve the problem by (effectively) re-defining varx within the expression i.e. varx = C wrote: > > It depends. The loop is general-purpose solution and works always. > > In some cases loops can be eliminated using Cum(), ValueWhen(), > AMA, AMA2. > > Best regards, > Tomasz Jane