XProfan

- Page 1 -

Nico
Madysa

so far I know, can Arrays not on NProcs transfer go. only reaches#, Strings$, Longs& and Floats! are allows.

XProfan X2
Win7 Home Premium, SP1, AMD Phenom(tm) II N830 Triple-Core Processor
Nico Madysa

Please see Datentypen: [...]

Nativ go (How of Nico already erwähnt) Strings, Floats & Longs supported.

If you a Wertekette How Arrays transfer want then fetch you simply

a memory z.B. by globalAlloc:
Compile Mark Separation

var meinSpeicher&=globalAlloc(gPTR,20)// 20 Byte Speicher reservieren und Handle in meinSpeicher& merken

and lay there into your values:
Compile Mark Separation

long meinSpeicher&,0=10,20,30,40,50

One Long needed in memory 4 byte and a Float 8 -

5 Floats into memory lay means so:
Compile Mark Separation

//speicher holen
var meineFloats&=globalAlloc(gPTR,5*8)
//floatwerte in den Speicher legen
float meineFloats&,0=1.1,2.2,3.3,4.4,5.5
//speicher an nProc übergeben
print myNProc(meineFloats&)
//speicher freigeben
globalFree(meineFloats&)
waitinput
end

nProc myNProc

    parameters meineFloats&
    //irgendwas mit floats machen... geben wir einfach z.B. den 4. Floatwert zurück
    return getFloat(meineFloats,8*3)

XProfan X2
Win7 Enterprise, SP1, Xeon X5670 @ 2.93GHz

Andreas
Koch

thanks for response.

then have I iFs Posting here well something wrong interprets.

[...]

whom Posting very?

Nico
Madysa

Presumably whom here: [...]

because it therein so sound, as should one only the Schlüsselwort "proc" through "nproc" supplant and everything would knorke.

XProfan X2
Win7 Home Premium, SP1, AMD Phenom(tm) II N830 Triple-Core Processor
Nico Madysa

XProfan X2
Win7 Enterprise, SP1, Xeon X5670 @ 2.93GHz

Andreas
Koch

open stood, had You only in keinem your Posts written, that with the Arrays none goes. I had the so interprets, that it too with Arrays functions (and me already gefreut...). really had I it from the last Posting entnehmen can, me was by then only yet never noticed, that one on APIs no Arrays commit. but presumably wanted You me there very the communicating, what now To this Topic here led has.
Sorry for circumstances. be sometimes bisschen heavy of concept.

understand. the here once durchlesen: [...]

-

its quasi one very very simple Profan but even as Support To XProfan

with all positive properties one nativen Profans.

open stood, had You only in keinem your Posts written, that with the Arrays none goes.

Yes, simply because the "soooo" indeed not integrally is correct, because even if The aray-

spelling not is used, so are Arrays Yes Datenketten and these against

are slight by memory anzufertigen and against too slight with nProcs

verwertbar. I had attempts for 1-2 Examples To give. If you not weiterkommst

then simply fix report because then basteln we the together.

funkheld

Error Message:
meineFloats undeklariert....

greeting

XProfan 11

certainly because I one Undzeichen forget have:

getFloat(meineFloats&,8*3)

How written simply into editor here typed

XProfan X2
Win7 Enterprise, SP1, Xeon X5670 @ 2.93GHz

Andreas
Koch

Sorry people, i'm again steckengeblieben.
now have I The Arrays![] in reaches# transformed and the transfer, so far so well. it shining too with Procs correctly. To calculate.
but with nProcs beschwert it itself in the procedure DreiecksmatrixTest first, that FHG% and Anzahl_Lastfaelle% not declared seien (though The still with transfer go). If I The deklariere, I get subsequently a seltsame Error Message, see Appendix.
Compile Mark Separation

Window 800,800
var Size%=6
Declare A#, Zeile&, Spalte&, Diagonale&, Faktor!, aux!,L#,Lastfall%,V#
DIM A#,Size%*Size%*8
DIM L#,Size%*8
DIM V#,Size%*8
Lastfall%=0
'Erzeugung zweier Zufallsmatrizen

WhileLoop 0,Size%-1

    Spalte&=&loop

    WhileLoop 0,Size%-1

        Zeile&=&loop
        SMAT(A#,Size%,Spalte&,Zeile&,1+Rnd())

        If Spalte&<1

            SMAT(L#,Size%,Spalte&,Zeile&,1+Rnd())

        EndIf

    EndWhile

EndWhile

'Umformung, Erzeugung einer Dreiecksmatrix

nproc DreiecksmatrixTest

    Parameters A#, L#, FHG%, Anzahl_Lastfaelle%
    Declare Diagonale&,Zeile&,Lastfall&,Faktor!,Spalte&, aux!, FHG%, Anzahl_Lastfaelle%

    WhileLoop 0,(FHG%-1),1

        Diagonale&=&loop

        WhileLoop Diagonale&+1,FHG%-1,1

            Zeile&=&loop
            '      WaitInput
            Faktor!=LMAT(A#,FHG%,Diagonale&,Zeile&)/LMAT(A#,FHG%,Diagonale&,Diagonale&)

            if Faktor!<>0

                WhileLoop 0,Anzahl_Lastfaelle%-1

                    Lastfall&=&loop
                    aux!=LMAT(L#,FHG%,Lastfall&,Zeile&)-Faktor!*LMAT(L#,FHG%,Lastfall&,Diagonale&)
                    SMAT(L#,FHG%,Lastfall&,Zeile&,aux!)

                EndWhile

                WhileLoop 0,FHG%-1,1

                    Spalte&=&loop
                    aux!=LMAT(A#,FHG%,Spalte&,Zeile&)-Faktor!*LMAT(A#,FHG%,Spalte&,Diagonale&)
                    SMAT(A#,FHG%,Spalte&,Zeile&,aux!)

                EndWhile

            EndIf

        EndWhile

    EndWhile

EndProc

nProc LMAT

    Parameters Bereich#, Size%, Spalte& , Zeile&
    Declare Result!
    Result!=Float(Bereich#, (Spalte&*Size%+Zeile&)*8)
    '  Print "Spalte&",Spalte&,"Zeile&",Zeile&, ": ", Result!
    Return Result!

EndProc

nProc SMAT

    Parameters Bereich#, Size%, Spalte&, Zeile&,Wert!
    ' Locate 1,1
    ' Print "Spalte&",Spalte&,"Zeile&",Zeile&,": ",Wert!
    Float Bereich#,(Spalte&*Size%+Zeile&)*8=Wert!

EndProc

'Rücksubstitution

nProc SolverTest

    Parameters Lastfall&, A#,L#, V#, FHG%
    Declare aux!, Zeile&, Spalte&, Diagonale&

    WhileLoop FHG%,1,-1

        Diagonale&=&loop-1
        aux!=0

        WhileLoop Diagonale&+1,FHG%-1,1

            Spalte&=&loop
            aux!=aux!+LMAT(A#,FHG%,Spalte&,Diagonale&)*LMAT(V#,FHG%,Lastfall&,Spalte&)

        EndWhile

        aux!=(LMAT(L#,FHG%,Lastfall&,Diagonale&)-aux!)/LMAT(A#,FHG%,Diagonale&,Diagonale&)
        SMAT(V#,FHG%,Lastfall&,Diagonale&,aux!)

    EndWhile

EndProc

DreiecksmatrixTest(A#, L#, Size%, 0)
SolverTest(0,A#,L#, V#, Size%)

WhileLoop 0,Size%-1

    Print LMAT(V#, Size%,0, &loop)

EndWhile

WaitInput
Dispose L#, V#, A#

have a little bit patience with me, I work not yet such a long time with Profan....

bild2.jp2	167 kB

Hochgeladen:	08/31/12

Downloadcounter:
		Download

08/31/12 ▲

No worry, there's In any drop sufficient patience present.

You verwendest into nProcs Integer ( % ) -

in nProcs there Longs ( & ), Floats ( ! ) and Strings ( $ ) -

use means instead of the Integer ( % ) The Longs ( & ).

The nProcs are manchen Things something pingeliger -

so must Float-Konstanten always with Dezimaltrenner/ point

ausgeschrieben go.

Z.B. 1.0 instead of 1 or 0.0 instead of 0.

You write: if factor!<>0 instead of if factor!<>0.0

therefore bleats XPSE the Long given is ( The 0 ) but Float expects becomes

in the function dreiecksmatrixtest.

You write in the nProc SMAT: Parameters area#, should but area&

heissen.

hint: but not Float area#,(Spalte&*Size%+Zeile&)*8=Wert! what but
Float area&,(Spalte&*Size%+Zeile&)*8=Wert! heissen should in nProcs rather
setFloat(Bereich&,(Spalte&*Size%+Zeile&)*8,worth!) write.

too into others Procs The Memory-Variables instead of with # with & write.

08/31/12 ▲

XProfan X2
Win7 Enterprise, SP1, Xeon X5670 @ 2.93GHz

- Page 2 -

Andreas
Koch

thanks for Help and the Lehrstunde, I faith, there had I dusselig sought. now runs it correct through what about me have what learnt.

whether it calculate properly, must I yet Verify. on and itself ought to the thing but one lineares Gleichungssystem solve can (sofern on the Hauptdiagonalen no nobodies come along).
Compile Mark Separation

Window 800,800
var Size&=6
Declare A#, Zeile&, Spalte&, Diagonale&, Faktor!, aux!,L#,Lastfall%,V#
DIM A#,Size&*Size&*8
DIM L#,Size&*8
DIM V#,Size&*8
Lastfall%=0
'Erzeugung zweier Zufallsmatrizen

WhileLoop 0,Size&-1

    Spalte&=&loop

    WhileLoop 0,Size&-1

        Zeile&=&loop
        SMAT(A#,Size&,Spalte&,Zeile&,1+Rnd())

        If Spalte&<1

            SMAT(L#,Size&,Spalte&,Zeile&,1+Rnd())

        EndIf

    EndWhile

EndWhile

'Umformung, Erzeugung einer Dreiecksmatrix

nproc DreiecksmatrixTest

    Parameters A&, L&, FHG&, Anzahl_Lastfaelle&
    Declare Diagonale&,Zeile&,Lastfall&,Faktor!,Spalte&, aux!

    WhileLoop 0,(FHG&-1),1

        Diagonale&=&loop

        WhileLoop Diagonale&+1,FHG&-1,1

            Zeile&=&loop
            '      WaitInput
            Faktor!=LMAT(A&,FHG&,Diagonale&,Zeile&)/LMAT(A&,FHG&,Diagonale&,Diagonale&)

            if Faktor!<>0.0

                WhileLoop 0,Anzahl_Lastfaelle&-1

                    Lastfall&=&loop
                    aux!=LMAT(L&,FHG&,Lastfall&,Zeile&)-Faktor!*LMAT(L&,FHG&,Lastfall&,Diagonale&)
                    SMAT(L&,FHG&,Lastfall&,Zeile&,aux!)

                EndWhile

                WhileLoop 0,FHG&-1,1

                    Spalte&=&loop
                    aux!=LMAT(A&,FHG&,Spalte&,Zeile&)-Faktor!*LMAT(A&,FHG&,Spalte&,Diagonale&)
                    SMAT(A&,FHG&,Spalte&,Zeile&,aux!)

                EndWhile

            EndIf

        EndWhile

    EndWhile

EndProc

nProc LMAT

    Parameters Bereich&, Size&, Spalte& , Zeile&
    Declare Result!
    Result!=Float(Bereich&, (Spalte&*Size&+Zeile&)*8)
    '  Print "Spalte&",Spalte&,"Zeile&",Zeile&, ": ", Result!
    Return Result!

EndProc

nProc SMAT

    Parameters Bereich&, Size&, Spalte&, Zeile&,Wert!
    ' Locate 1,1
    ' Print "Spalte&",Spalte&,"Zeile&",Zeile&,": ",Wert!
    Setfloat(Bereich&,(Spalte&*Size&+Zeile&)*8,Wert!)

EndProc

'Rücksubstitution

nProc SolverTest

    Parameters Lastfall&, A&,L&, V&, FHG&
    Declare aux!, Zeile&, Spalte&, Diagonale&

    WhileLoop FHG&,1,-1

        Diagonale&=&loop-1
        aux!=0

        WhileLoop Diagonale&+1,FHG&-1,1

            Spalte&=&loop
            aux!=aux!+LMAT(A&,FHG&,Spalte&,Diagonale&)*LMAT(V&,FHG&,Lastfall&,Spalte&)

        EndWhile

        aux!=(LMAT(L&,FHG&,Lastfall&,Diagonale&)-aux!)/LMAT(A&,FHG&,Diagonale&,Diagonale&)
        SMAT(V&,FHG&,Lastfall&,Diagonale&,aux!)

    EndWhile

EndProc

DreiecksmatrixTest(A#, L#, Size&, 0)
SolverTest(0,A#,L#, V#, Size&)

WhileLoop 0,Size&-1

    Print LMAT(V#, Size&,0, &loop)

EndWhile

WaitInput
Dispose L#, V#, A#

Nachtrag: The Source is meanwhile examined and it shining correctly To calculate. he's in the doing very plenty faster as before. I have no detailed Messung made, but factor 2000 is realistic.

08/31/12 ▲