[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[amibroker] Re: Subject: Results of Scan/Exploration



PureBytes Links

Trading Reference Links

Explore ^NDX for the n=1 last bar with
H2=HHV(IIf(H!=H1,H,0),100);
H3=HHV(IIf(H!=H1 AND H!=H2,H,0),100);
H4=HHV(IIf(H!=H1 AND H!=H2 AND H!=H3,H,0),100);
H5=HHV(IIf(H!=H1 AND H!=H2 AND H!=H3 AND H!=H4,H,0),100);
Filter=1;
AddColumn(H1,"h1");
AddColumn(H2,"h2");
AddColumn(H3,"h3");
AddColumn(H4,"h4");
AddColumn(H5,"h5");
The result is the top5 H .
If we take the 100 results of the 
Filter=1;
AddColumn(Stochd(),"STOCHD");
exploration of n=1 last bar of the 100 N100 stocks and put them as 
the last 100 values of H of a certain stock, we could do it.
DT
--- In amibroker@xxxxxxxxxxxxxxx, "DIMITRIS TSOKAKIS" <TSOKAKIS@xxxx> 
wrote:
> Although it sounds simple, it isnīt.
> It is also difficult to find the top5 H of the last 100 bars of 
MSFT.
> [this is the horizontal sorting, in exploration result we need the 
> vertical one...
> DT
> --- In amibroker@xxxxxxxxxxxxxxx, uenal.mutlu@xxxx wrote:
> > Hi DT,
> > yes such a function in AFL would indeed be nice.
> > Here is my AFL suggestion:
> > 
> >   CutTable(columnNbr, n, bAscending = true);
> > 
> > This would be evaluated after the the result table 
> > was generated and would sort it and cut after the first
> > n records.
> > 
> > Another method would be using OLE/COM interface by executing
> > the users AFL script and exporting the table programmatically, 
then
> > sorting and cutting within the other program or AFL plugin DLL, 
and
> > putting the remaining tickers into a watchlist and running the 
users
> > script a second time but now on this watchlist... Hmmm.. maybe not
> > very efficient. But generally, this is doable but tedious. The 
best 
> would 
> > be the proposed new AFL method above.
> > 
> > UM
> > 
> > 
> > ----- Original Message ----- 
> > From: "DIMITRIS TSOKAKIS" <TSOKAKIS@xxxx>
> > To: <amibroker@xxxxxxxxxxxxxxx>
> > Sent: Wednesday, April 23, 2003 9:14 AM
> > Subject: [amibroker] Re: Subject: Results of Scan/Exploration
> > 
> > 
> > > Herman,
> > > Stewart asked at
> > > http://groups.yahoo.com/group/amibroker/message/38506
> > > something simple : how to select the top5 of an exploration. 
Any 
> idea 
> > > for this request ?
> > > [If I understood well, to run an exploration
> > > Filter=1;
> > > AddColumn(MACD(),"MACD");
> > >  for 100 stocks and see in the result list ONLY the top5 MACDs, 
> > > nothing else]
> > > I do not see how can I do it. 
> > > [I hope to avoid MAX(Foreign("~AAPL-MACD","C"),MAX(Foreign
("~ABGX-
> > > MACD","C"),...]
> > > Any idea appreciated.
> > > Dimitris Tsokakis
> > > --- In amibroker@xxxxxxxxxxxxxxx, "Herman vandenBergen" 
> <psytek@xxxx> 
> > > wrote:
> > > > Using array subscripts you can put any kind of data into a 
> array:
> > > >     Data[0] = YourParameter1;
> > > >     Data[1] = Your{arameter2;
> > > > YourParameter could also use subscripts...
> > > > 
> > > > When you want to save the data into a stock-specific array 
you 
> use 
> > > the Atc,
> > > > some thing like this for the Explorer:
> > > > 
> > > > AddToComposite(Data,"~"+Name()+"-Data","X",1|2|4|16); // Use 
in 
> > > Explorer,
> > > > Data here refers to the array you filed using subscripts. 
This 
> > > array will be
> > > > of length equal to number of bars in the current stock - a 
bit 
> of
> > > > over-kill - but this means you can put a ton of data into a 
> > > Composite (5
> > > > fields: OHLCV). The great thing is that this information 
> remains on 
> > > Disk for
> > > > re-use at any time from any program, use Foreign("~"+Name()+"-
> > > Data","X");
> > > > here X can be any of OHLCV.
> > > > 
> > > > You can save system calibration data for automatic recall. 
You 
> > > cannot save
> > > > text in a Composite however you can save Status("StockNum"), 
> this 
> > > is an
> > > > ordinal number pointing to your stock in your group (it 
changes 
> as 
> > > you
> > > > change groups!!).
> > > > 
> > > > Best regards,
> > > > Herman
> > > > 
> > > > 
> > > > 
> > > > 
> > > >   -----Original Message-----
> > > >   From: Stewart [mailto:stewart@x...]
> > > >   Sent: April 22, 2003 10:29 AM
> > > >   To: amibroker@xxxxxxxxxxxxxxx
> > > >   Subject: Re: [amibroker] Subject: Results of 
Scan/Exploration
> > > > 
> > > > 
> > > >   but using AddtoComposite(), is there a way to relate "a 
row" 
> to a 
> > > specific
> > > > ticker?
> > > >     ----- Original Message -----
> > > >     From: Dimitris Tsokakis
> > > >     To: amibroker@xxxxxxxxxxxxxxx
> > > >     Sent: Tuesday, April 22, 2003 5:56 PM
> > > >     Subject: [amibroker] Subject: Results of Scan/Exploration
> > > > 
> > > > 
> > > >     Suppose we run an exploration for 100 stocks
> > > >     X=StochD();
> > > >     Filter=1;
> > > >     AddColumn(X,"StochD");
> > > >     for one day.
> > > >     The result is an 100-dimension vector.
> > > >     We can save only in 6 fields through AddToComposite() 
> function, 
> > > namely
> > > > C, O, H, L, V, I.
> > > >     If we place the first 6 results to each field, we have no 
> place 
> > > to save
> > > > the rest 94 results.
> > > >     It is a 100X100 diagonal matrix, as in the att. gif
> > > >     but, even if we could, the result would not be an array.
> > > >     An array has one numerical value per day. In this case we 
> would 
> > > have a
> > > > set of 100 numerical values per day
> > > >     and we would create an 100-dimensional "Hyper array".
> > > >     It needs specific imagination to understand the use of 
this 
> > > creature.
> > > >     DT
> > > > 
> > > >     There's no way to store the results of an Exploration to 
an 
> > > array,
> > > > right?
> > > > 
> > > >     Thanks,
> > > > 
> > > >     Stewart


------------------------ Yahoo! Groups Sponsor ---------------------~-->
DVD Rentals with No Late Fees - Try Netflix for FREE!
http://us.click.yahoo.com/ZKLNcC/pEZFAA/AG3JAA/GHeqlB/TM
---------------------------------------------------------------------~->

Send BUG REPORTS to bugs@xxxxxxxxxxxxx
Send SUGGESTIONS to suggest@xxxxxxxxxxxxx
-----------------------------------------
Post AmiQuote-related messages ONLY to: amiquote@xxxxxxxxxxxxxxx 
(Web page: http://groups.yahoo.com/group/amiquote/messages/)
--------------------------------------------
Check group FAQ at: http://groups.yahoo.com/group/amibroker/files/groupfaq.html 

Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/