Mike, Code addition does seem to work. Looks like I did not save change before trying. The code excluded short history tickers from any kind of backtest. Now, I need to read documentation again to see if way to identify rather than just exclude. Thanks as I am getting closer.
--- In [email protected], "bistrader" <bistra...@...> wrote: > > Mike, > > Good idea. Tried via following and did not remove or otherwise identify > short history tickers. To answer your question, I understand how to do with > 2 step process, but this is not at all productive. I want to run a number of > different range periods via individual backtest on watchlist. I say, let me > look at this period as these types of securities did not do well in this > period and I want to see if Rsi addition adds some value. Then I go on to > look at other periods that these tickers did not do well in to see if I have > something. Then I say lets now see what happens for the longer period of > time when these types of securities were trending up as well as down and see > what happens with Rsi addition. Now, during these different periods, > AmiBroker provides the stats, but one has NO idea if these stats are for the > entire period analyzed or not as I have not been able via AmiBroker Support > or via my code changes to get AmiBroker to identify or exclude short history > tickers from the individual backtest analysis. > > I am thinking that there must be a way to get this. > > Oh, here is what I tried based on my understanding of your feedback. Did not > work. > > StartDate = BeginValue(BarIndex()); > //EndDate = EndValue(BarIndex()); > Range_Start = Status("rangefromdate"); > Symbol_Start = ValueWhen( BarIndex() == StartDate, DateNum(), 1); > diff = Range_Start - Symbol_Start; > //Close = IIf(diff < 0, Null, Close); > > if (LastValue(diff) < 0) > { > SetOption("GenerateReport", 0 ); // suppress generation of report > } > > Thanks. > > Bert > > Thanks > > Bert > > --- In [email protected], "Mike" <sfclimbers@> wrote: > > > > See if SetOption("GenerateReport", ...) can be used for your needs. The > > documentation reads as though it only applies during optimization, but it > > might work. > > > > Otherwise, is there a special reason why you don't want to run an initial > > exploration to filter out those symbols that don't satisfy your date > > requirements? > > > > Mike > > > > --- In [email protected], "bistrader" <bistrader@> wrote: > > > > > > I wrote to AB support, but have not heard from them on how I might change > > > afl to identify (ideal) or exclude tickers in a watchlist with short > > > history when running Individual Backtests > > > > > > I run individual basktests on a watchlist, but do not want to get stats > > > for tickers with history that does NOT go back to AA selected start date. > > > The only thing I can come up with is the following and this does not > > > work. Ideally, I want these tickers identified, but if not possible then > > > just to exclude them from the Individual Backtest report. > > > > > > I placed the following in afl, thinking that it would force no trades or > > > null for short history tickers in my watchlist. It does cause close to go > > > to null in exploration, but backtest still provides the stats for short > > > history tickers which is not what I want. I want the stats to be > > > identified (ideal) or excluded for those tickers that do NOT history back > > > to at least the AA start (called from) date. > > > > > > StartDate = BeginValue(BarIndex()); > > > //EndDate = EndValue(BarIndex()); > > > Range_Start = Status("rangefromdate"); > > > Symbol_Start = ValueWhen( BarIndex() == StartDate, datenum(), 1); > > > diff = Range_Start - Symbol_Start; > > > Close = IIf(diff < 0, Null, Close); > > > > > > Direction appreciated on how can handle via change in afl code. > > > > > > Thanks > > > > > > Bert > > > > > >
