☰

» »

Digitalna evolucija

Digitalna evolucija

««
22 / 29
»»

romci ::

snow: true. res je izboljsalo algo, ki sem ga pokvaru - ampak ga je zgleda izboljsalo se naprej do te mere, da je bil se boljsi od originala (ne samo od "pokvarjenga").

Tisto da se pred zanko nabira racunanje je pa tako kot je Thomas reku - to se zracuna samo enkrat, medtem ko se same zanke izvedejo XYZ krat, in ni nekega profita od optimizacije solate.

Res pa lahka poskusmo optimizirat po segmentih. Recimo das samo najbolj notranjo zanko med $BES in $EES. Al pa notranji dve.
Ampak osebno mislim da bi dal s tem le manjsi margin za izboljsave.

LP,
Roman
-- not all those who wander are lost...

Jst ::

Torej... imam service, ki sam po sebi ne dela nič, ampak skrbi za delovanje zunanjega programa. V tem primeru sort questa. Imam pa še Control Center, ki skrbi za ta service. V tem control centru nastaviš kateri zunanji program naj teče. Določiš mu tudi ali je main window visible ali ne.

Prvo sem naredil en testni programček, ki v eno memo polje vrže dve random številke in potem dve random vrstice pomnoži in zapiše spet v memo. rekurzija.

Ta testni programček sem dal v service in je lepo delal. Kadar je delal v ozadju in ko sem hotel main window visible je delalo. Ko sem spreminjal prioriteto, je zadeva delovala. Tudi ko sem hotel program končati, se je končal brez napak.

Potem sem pa v ta service vrgel sort quest. Spreminjanje prioritete je delovalo. Tudi ko sem main window skil je še delalo. Ko sem pa hotel priklicati main window, pa bum. Zadeva crkne. CPU Load gre na 0% in main window se ne prikaže. Problem je v Sort Questu.

Any ideas?

Ko to poštimam, lahko postam link, če ima kdo željo, da bi uporabljal. Seveda bi s tem lahko laufali tudi druge programe.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Thomas ::

Hm .. študiramo, kaj bi bilo ...

OwcA ::

Ne upam si dati roke v ogenj, ampak zdi se mi, da ima tole čud nekje malo pušča.
Otroška radovednost - gonilo napredka.

Thomas ::

Memory leakeage? Possible.

Zgodovina sprememb…

  • spremenil: Thomas ()

snow ::

Jst jaz bi probal mal ta tvoj control service.
Tist main window me niti nebi motil nekaj preveč.
Random mutation plus nonrandom cumulative natural selection - Richard Dawkins

Thomas ::

Huh ja, danes je bil urnik dan, Zeitplantag. :))

Tko da tale konec je nekoliko stal.

Kar se tiče nestabilnosti kode ... dopuščam da pušča in zdej ravno teče en tozadeven test. Bomo videli kako in kaj, to bo že.

Jst ::

Bom dal service program, koj ko gre čez bug test in se naredi lepši GUI.


...

Thomas:

1. Če ne inštaliraš v default dir, potem pri zagonu crkne.

2. Piše output: "C:\Program Files\Critticall_Sort_Quest\Samples\fs.c_c", vendar le ta ne obstaja. Zakaj to??

3. Ko daš File/Exit in ponovno zaženeš, ne nadaljuje tam, kjer je prej končal. (Oziroma, kako to vidim?)



....


Kaj pomeni Pausing ON/OFF, $Pausing?

Prvič, ko sem pustil nekaj več kot 15 ur, mi je pisalo, 41.9% of computing originally needed.


Ali mi lahko malo razložiš, kaj pomenijo stvari, ki so na zaslonu. Razumem samo kodo. :)
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Thomas ::

> 1. Če ne inštaliraš v default dir, potem pri zagonu crkne.

Tako je. Background dependant je. Če mi poveš en dober razlog zakaj naj bi ne bil - ne bo.

> 2. Piše output: "C:\Program Files\Critticall_Sort_Quest\Samples\fs.c_c", vendar le ta ne obstaja. Zakaj to??

"C:\Program Files\Critticall_Sort_Quest\Samples\fs.c_e"

> 3. Ko daš File/Exit in ponovno zaženeš, ne nadaljuje tam, kjer je prej končal. (Oziroma, kako to vidim?)

Ne, vedno gre znova raziskovat nove veje. Tako je optimalno, sicer bomo čemeli v N evolucijskih mrtvih rokavih. Tako pa M*N mrtvih rokavih. Večja je verjetnost, da bo eden od njih bolj optimalen, kot če jih je samo N.

> Kaj pomeni Pausing ON/OFF, $Pausing?

Pausing after every enhancement. By default it is OFF. $Pausing means - as declared in file.

> Prvič, ko sem pustil nekaj več kot 15 ur, mi je pisalo, 41.9% of computing originally needed.

To je pomenilo, da skoraj 60% od Fat Bubble Shell sort korakov je zmetal proč. Fat bubble je invencija iz Several Unique. Tukaj aplicirana na Shell.

> Ali mi lahko malo razložiš, kaj pomenijo stvari, ki so na zaslonu. Razumem samo kodo. :)

Next post, zdej mam večerjo. :)

Thomas ::



Pojasnila k ekranu.

Zgodovina sprememb…

  • zavarovalo slike: OwcA ()

Jst ::

> > 1. Če ne inštaliraš v default dir, potem pri zagonu crkne.

> Tako je. Background dependant je. Če mi poveš en dober razlog zakaj naj bi ne bil - ne bo.

Potem pa ne rabiš dati izbire direktorija inštalacije.

btw: Kaj misliš s tem, background dependant?


> > 3. Ko daš File/Exit in ponovno zaženeš, ne nadaljuje tam, kjer je prej končal. (Oziroma, kako to vidim?)

> Ne, vedno gre znova raziskovat nove veje. Tako je optimalno, sicer bomo čemeli v N evolucijskih mrtvih rokavih. Tako pa M*N mrtvih rokavih. Večja je verjetnost, da bo eden od njih bolj optimalen, kot če jih je samo N.


Ne razumem najbolj. A si zapomne, kaj je prej delal? In po ponovnem zagonu to upošteva ?


....


Aja, zakaj mi pa samo 47% CPU kuri?

Procesor je pa p4 HT.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

snow ::

Glede 47%
Zato ker je procesor HT(navidezna 2 procesorja) ti kuri sam 'pol'.
Zaženi 2x.

> Ne razumem najbolj. A si zapomne, kaj je prej delal? In po ponovnem zagonu to upošteva ?

Niti ne, ampak zadeva je dost random narejena, da bo zelo zelo redko šel po isti poti. Kot evolucija... če bi se na začetku življenja mal drugače formirale molekule, bi bil svet danes popolnoma drugačen. Ista situacija je tukaj.
Random mutation plus nonrandom cumulative natural selection - Richard Dawkins

Jst ::

> ker je procesor HT(navidezna 2 procesorja) ti kuri sam 'pol'

Torej je to zelo slabo optimizirano za novejše CPUje.


> > A si zapomne, kaj je prej delal?

> Niti ne

Kakšen Quest je potem to, če vsakič začne od začetka (istega sourcea)?

Hm...
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

OwcA ::

Meni se vedno ena instanca sesuje, če nekaj časa poganjam dve skupaj. Kar niti ni tako zelo čudno, glede na to, da zgleda, da obe pacata po istih datotekah.
Otroška radovednost - gonilo napredka.

Jst ::

Priporočam ogled Intelovih ppsjev iz letošnjega seminarja o programiranju za HT.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

snow ::

Jst daš link za lenobo? :) :D
Random mutation plus nonrandom cumulative natural selection - Richard Dawkins

Thomas ::

Ja, stepeta se zaradi fajlov. Critticall se ne, ta dva pa ... thnx ...

....

Jst ::

snow: V novicai na st je bil link. Nekaj v stilu meltdown.zip. Kjer je bilo veliko pps-jev.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Jst ::

Zmotil sem se:

microsoft.com
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

snow ::

Hvala!
Random mutation plus nonrandom cumulative natural selection - Richard Dawkins

Jst ::

Če mi bo uspelo "stabilizirati" moj service z delom z sortquestom, bom čez vikend (če bom imel možnost - velika verjetnost da bom) dal laufati na en server (dual xeon, windows 2003 advanced server).


Zanima me pa naslednje:
Na kakšne načine lahko kontroliraš izvajanje poljubnega programa? Jaz bi rabil eno takšeno freeze možnost. Da, ko bi kakšna druga aplikacija kurila več kot 10% CPUja, bi ta service freeznil zunanji program, ki ga kontrolira. Nekako v stilu borland delphi/c bouilder, kjer imaš "Pause" opcijo.


Thomas, glej zasebna sporočila.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Jst ::

Sem testno (za 5 min) zalaufal ta sort quest na enemu dual xeon serverju na win2003 advanced server. Tam je bila zgodba še bolj klavrna. CPU usage sort questa je bil okoli 23%, kar me je kar nekako šokiralo. In sem šel malo raziskovat. Ni bilo potrebno dosti dela: Ctrl-alt-del / task manager / view / select columns / thread count. Število threadov je večinoma 1, redko 2. To je tudi vzrok, zakaj ne izkoristi na HT in na dualcpu mašinah ves computing power, ki ga ima na voljo.


...

Glede run as service programa:
Še vedno pa študiram kaj je bolje. Ali da bi Crittical gnal kot service in z enim programom kontroliral njegovo izvajanje (prioriteta, bring to front main window, pauza). Ali pa da bi s servisom gnal katerikoli program v "debug mode-u", zagnan s CreateProcess Win32 API klicom.

Sem vprašal na delphi-si, kako bi izvedel prvi način...
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Jst ::

Glede drugega načina. V mislih imam eno tray ikonco, kjer bi bili preprosti ukazi: Stop, Pause, Continue, Bring To Front, Hide, Exit (tega kontrolerčka).

Ker nisem designer bi potreboval eno ikono. Če ima kdo 5 min časa naj mi nariše kakšno ikono za v tray. Vsi predlogi dobrodošli na mail v profilu.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Zgodovina sprememb…

  • spremenil: Jst ()

Thomas ::

Ne moreš laufat osnove tega v več kot v enem threadu. Lahko pa poženeš N neodvisnih Critticallov, ne pa SortQuestov.

TO bomo spremenili.

Jst ::

Aja mail je "matej7 AT siol.net". Ga nimam v profilu.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Jst ::

Seveda lahko laufaš v večih threadih.


Kolikor razumem koncept CR, stvar deluje tako (popravite me, če se motim):

1. Ima kodo, ki jo nekaj časa random obmetava.
2. Pogleda, če je sintaktično pravilna.
3. Primerja kodi in hitrejšo vzame.
4. Rekurzija na 1.

Te vse točke daš lahko v threade.

Kako pa?

Prvi thread: je thread, ki obmetava kodo. Ko jo neha obmetavati jo da drugemu threadu, da preveri, če je sintaktično pravilna in počaka na signal, da je drugi thread končal. Prvi thread potem naprej obmetava kodo, ker predpostavlja, da je prejšnji hitrejši. V tem času pa zalaufaš tretji thread, ki primerja kodo, ki jo je vrgel prvi thread.

Prvi thread NE sme končati, če je tretji thread aktiven.

Tukaj imaš že 3 threade.

Četrti thread pa skrbi za izpis na zaslonu.


...

Dodatna hitrost se ti pozna samo pri novejših HT procesorjih in dual (ali več) sistemih.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

snow ::

Kaj če bi pa dal vse te korake v več threadov?
Pa da pol glavni skrbi za display najboljšega...oni tam pa vsak po svoje mal.

zadevi se reče 'koevolucija'.
pa če maš še pametno komunikacijo(=prehajanje osebkov oz. genomov), lahko dost pridobiš če maš večprocesorski sistem ali HT.
Random mutation plus nonrandom cumulative natural selection - Richard Dawkins

Zgodovina sprememb…

  • spremenilo: snow ()

Jst ::

+ toliko kolikor je CPUjev (tudi virtualnih - kot pri p4 HT), tolikokrat več teh threadov laufaš.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Thomas ::

Sorting večjega števila elementov se da razporediti na threade, ja.

Množenje velikih števil pa samo teoretično. Ni haska, ker so vmesni rezultati esencialni za vsako naslednjo stopnjo računanja.

Tudi tukajle ne gre v resnici za plain evolucijo - mutacije in izvajanje mutiranih programov - pač pa bolj za spraševanje, kaj bi taka mutacija spremenila na outputu in hitrosti.

Lahko povem, da se le računa (transformira) neka velika matrika. Podobno množenju.

Da bi bil izpis pa svoj thread - to bi blo pa cool. Bo cool.

snow ::

vseeno mi še noben ni povedal zakaj nebi en program mogu na večih threadih laufat koevolucijo in bi en glavni thread gledal za najboljšim primerom, pa še izpisoval zraven?
tak več 'worker' threadov pa en glavni...
Random mutation plus nonrandom cumulative natural selection - Richard Dawkins

Thomas ::

To bi se pa dalo. To pa definitivno.

Da sprožiš več teh procesov. Vsakega v svojem threadu, potem pa en thread išče prvaka. To pa.

AMPAK. Je čisto ista figa, če poženeš 4 Critticalle (4 SortQueste še ne gre) in vsak telovadi svoje. Prvaka pa vidiš po nekem času. Zgodnji favorit pa ni nujno prvak.

Thomas ::

In ker zgodnji favorit ni nujno prvak, je mediiranje petega threada katerega bo killal, zguba časa.

Besides, še hujše. Raziskovalni prostor je velik in mnogo poti zaželjenih. Prisilno kanaliziranje v isto strugo pa zmanjša verjetnost uspeha. Zakaj bi vsi delali na neki dosedaj najboljši varianti? Tisto bo podelal oni sam, če ni zataknjena. Če je, pa še drugih 10 tudi ne. Ti naj gredo zato drugam, po drugih poteh.

snow ::

saj ne rabijo delat isto. vsak po svoje. vsakemu svoj random generator, da dela svojo pot.
pol pa en glavni gleda.

ja lahko zalaufam dva critticalla, pa vsakemu pot štimam.. sam zakaj če lahko enega, pa ta dela dejansko na polno in daje ven najboljše kar se je spomnil katerikoli worker.
moram res tam klikat 2x pa štimat fajle :)



jaz bi to click and go away... ne pa click poišči pot, poišči tisto, click, poišči file, določi izpis... :\

aja ko smo že glih pri tem click... predlagam da bi se tisti _e dodajali imenu če je končnica že c_e
da pač sam kliknem :D
Random mutation plus nonrandom cumulative natural selection - Richard Dawkins

Thomas ::

Mnja ... maš point.

Ampak nadzorni thread pa je odveč. Ampak da so kar popakirani in delajo iz enega klika - to se pa moram strinjat.

Komunikacija med njimi, skozi nadzorni thread, je pa tako, kot bi gledal evolucijo slonov in krokodilov in ju koordiniral.

Ne, naj cveti 100 cvetov!

njok ::

No, na disku so se cudezno sprostili 3GB prostora. Tako je po dveh dneh evolucije rezultat sledec:

// The algorithm has been enhanced For 54.851%

 $DECLAREINT n h v j tt acritticall1 critticall2 

 $DIMENSION a[100]
 $INVAR n(100) a[](47,-30,2,-37,38,-48,-2,46,-18,27,19,38,-20,12,33,17,-3,47,-28,46,5,48,-37,-5,50,-31,19,17,-13,-43,32,-8,-4,30,7,-4,-41,45,14,40,-37,29,48,-5,-33,-26,40,3,34,36,41,23,27,-27,-24,-21,26,31,-45,1,-10,-46,-40,-25,-5,-31,-28,-47,17,-19,-8,-30,30,-20,-25,-13,-46,48,-19,-46,-41,-2,6,48,18,9,-46,-50,-38,25,-31,-46,-14,28,16,21,-9,20,-2,24)
 $INVAR n(100) a[](-9,-42,-41,-49,-8,29,-27,-35,29,-37,49,15,37,32,-1,-31,-38,-43,43,22,-39,39,39,-36,40,-14,-19,48,12,-2,-1,-34,16,33,30,43,-12,-47,-22,-1,-19,-41,-28,26,4,19,-33,6,-8,-24,-26,-50,50,-6,8,36,17,47,-17,-33,-42,1,3,10,-24,17,6,-50,-16,-32,-9,-21,3,-18,23,22,-23,25,-29,49,-5,-16,-28,41,-21,45,-16,-6,-15,36,-14,-12,10,11,-48,-44,10,-18,-19,26)
 $INVAR n(100) a[](-30,17,-23,18,-12,-14,7,35,49,-5,-16,-44,9,-46,49,40,31,-23,40,-14,14,-14,-7,15,21,-10,5,-26,3,-44,38,-37,-34,-21,-36,46,32,34,-27,-10,-14,-40,-4,-40,21,21,6,43,10,12,10,-22,-26,39,45,45,24,-14,30,5,-2,29,47,6,16,-35,46,15,30,-6,10,9,35,35,9,-40,-27,-26,34,-13,-50,1,-5,-42,38,43,25,-27,27,29,13,-42,-16,41,-34,40,-9,20,-22,-2)
 $INVAR n(100) a[](-21,46,41,10,16,14,40,29,-43,-38,11,-42,13,30,-42,33,-23,-47,-49,-10,15,30,-49,16,5,-4,47,-3,-12,18,-17,-14,-18,-27,33,4,30,-25,19,5,17,-9,7,-14,-19,17,29,29,-30,37,-14,-21,-50,-17,-41,30,-45,-45,25,-45,34,-29,46,-7,38,-30,-13,-2,49,-18,-10,1,48,26,-22,46,-32,12,-1,-5,-8,-37,26,20,-31,-32,-37,21,-19,36,29,37,-23,21,14,-49,50,38,50,-4)
 $INVAR n(100) a[](25,-23,15,0,-4,-14,-28,-43,15,0,23,0,23,15,9,-10,-12,46,5,-20,-20,13,-26,10,-3,42,3,15,21,-15,1,-4,-13,-12,-36,-48,-39,-47,24,-14,-14,-3,23,31,41,-22,13,-42,-3,-42,37,-28,1,21,-33,-13,-4,43,-13,-43,-41,30,-37,-10,39,25,26,-18,-8,-35,-42,44,-26,-48,-19,-2,29,50,-49,-9,-35,33,31,29,46,50,0,29,6,-37,11,-28,-30,-6,-4,-33,-39,-20,0,-19)
 $RINVAR a[](-50,50) n(1,2000) 
 $RETVAR a[] 
 $WEIGHTS lines=1 commands=1

// int a[100]; int n=0;int h=0;int v=0;int j=0;int tt=0;int acritticall1=0;int critticall2=0;

acritticall1=100;
if (n>acritticall1) {
    n=100;
}
$BES
critticall2=4;
v=critticall2<<v;
critticall2+=3;
critticall2=v+critticall2;
while (h!=critticall2) {
    j=critticall2;
    while (j<n) {
        tt=j-critticall2;
        v=a[j];
        tt=a[tt];
        if (tt<=v) {
            goto labelcritticall15;
        }
        acritticall1=j;
        while (critticall2!=n) {
            a[acritticall1]=tt;
            acritticall1=acritticall1-critticall2;
            if (acritticall1<critticall2) {
                break;
            }
            tt=acritticall1-critticall2;
            tt=a[tt];
            if (tt<=v) {
                goto labelcritticall13;
            }
        }
        labelcritticall13:;
        a[acritticall1]=v;
        labelcritticall15:;
        j+=1;
    }
    critticall2>>=2;
}
$EES
//$RINVAR  a[](-50,50) n(1,2000) 
$INVAR a[](40,29,8,34,17,42,49,34,-21,8,-33,-27,0,-40,-16,-49,-37,14,-46,8,-22,25,-7,-27,21,-30,22,8,20,39,-40,-26,28,37,14,22,32,-41,-43,-14,27,-28,-11,9,31,-3,11,46,18,14,-20,31,10,10,0,-17,37,47,-16,2,-24,12,-14,-25,9,-7,-15,-7,-27,-16,-24,-47,-12,3,1,-9,-13,-5,14,-31,-22,45,-28,19,28,21,4,11,31,18,-26,3,1,-15,-30,-18,27,-48,-36,10)  n(1947)  
$INVAR a[](-20,-27,29,-38,21,11,-17,-2,-41,15,2,-4,-29,19,11,0,28,-31,24,13,-5,30,50,-38,-47,9,39,21,-1,43,-49,6,-46,32,8,24,38,-49,-50,-1,-48,-47,4,22,-18,27,-47,10,22,18,42,-46,-23,-6,-35,7,6,10,40,15,-16,27,34,26,27,-29,13,12,-5,46,37,-18,-37,27,-2,-11,36,15,-1,45,24,14,8,2,47,-41,34,-34,-11,25,17,50,47,-26,23,-31,42,43,-18,-35)  n(544)  
$INVAR a[](-4,-22,34,12,-31,42,-26,-14,-3,-4,42,-11,-1,49,-21,-2,37,28,48,43,3,-15,-25,47,-38,-18,49,23,-7,30,19,0,48,21,-42,-45,-36,-2,-4,-33,-18,36,7,27,29,23,42,-36,6,-13,21,-20,-17,-7,28,-11,-36,-4,-22,15,8,50,44,-3,-3,-10,-42,-9,-1,38,50,-16,-39,9,-47,43,28,12,16,-20,-20,-33,-18,-3,34,1,43,44,33,-31,-32,-2,49,32,-39,-36,-41,-18,10,39)  n(87)  
$INVAR a[](-47,41,2,29,-10,-38,-45,-44,-43,-10,-34,20,-12,1,40,37,-10,-37,30,9,9,15,-29,-5,-44,-15,17,34,27,35,26,-28,41,-46,36,-23,-17,15,19,20,30,-26,15,-37,-22,-7,-23,-1,23,-14,-24,10,48,49,-43,43,19,6,38,42,11,-23,-13,-39,-15,-24,44,-45,11,-40,37,23,7,18,33,46,26,50,27,-30,22,-15,33,-18,-29,49,-38,10,33,38,-14,46,-26,15,6,-50,-2,6,10,19)  n(1727)  
$INVAR a[](-27,-8,-43,40,33,-46,-25,34,23,46,49,-6,46,-33,-13,-34,-17,-5,16,-43,-15,-13,27,11,-43,-46,47,32,25,18,-41,-21,-18,17,-48,10,-26,43,-3,-42,-44,-5,-28,-34,34,10,-5,18,-23,-10,21,29,-41,41,13,0,-24,-4,9,36,27,-24,28,-14,-29,31,-1,-34,9,45,1,-50,17,5,39,-40,-42,-26,11,-12,-39,11,34,15,-8,24,28,32,15,-42,-15,-39,-14,-46,-40,43,24,24,-18,28)  n(483)  
$INVAR a[](32,41,50,-33,-44,-26,39,49,38,-15,-1,-31,-4,-38,-34,-3,-43,48,-2,-38,11,-46,-25,-3,-45,14,14,-13,-17,15,-26,-42,-16,45,41,-20,-26,43,9,48,-24,35,12,-28,10,-10,45,9,-30,39,4,-16,-4,35,31,-1,-35,45,19,43,-38,10,-50,22,7,6,46,-6,41,-9,-49,-27,6,31,6,26,-27,35,-50,-9,-12,30,-23,-8,-47,-20,30,10,-30,12,3,-49,-7,3,5,25,-23,50,-23,6)  n(1591)  
$INVAR a[](16,-34,-28,3,11,-29,37,-33,-10,41,36,-13,-41,-17,17,16,-45,34,-44,19,-44,1,-18,30,-1,-8,-9,21,-25,-10,38,21,-31,-11,-11,49,-41,-2,-45,38,2,-25,46,-26,47,-42,4,-23,-16,-11,42,-20,13,-44,49,6,50,16,-38,49,-36,40,35,-20,-5,14,-1,-29,50,-5,-42,-19,44,-16,29,-44,1,-22,12,50,26,10,-20,42,19,-4,9,32,1,39,-26,22,-27,-46,3,27,5,34,-46,-32)  n(3)  
$INVAR a[](4,-45,-17,22,3,-22,-38,-34,15,-44,14,43,-45,-13,4,24,-23,-42,40,-4,-26,8,-36,-28,33,-20,15,-22,-26,-14,-40,-31,13,-27,-22,40,5,33,-46,29,-22,18,5,50,-42,49,49,-18,39,18,-8,11,-48,42,3,0,-49,-1,-29,34,-5,-36,-3,-34,5,27,-46,-37,38,49,11,15,-17,13,12,-31,-4,49,50,13,5,45,-17,-16,-33,-9,-16,12,-43,7,10,5,16,33,38,46,-7,19,14,45)  n(15)  
$INVAR a[](-24,17,44,-36,-25,-44,-36,-8,46,14,36,50,46,-46,-10,-46,-12,11,-22,3,-27,22,46,-47,-30,34,9,-16,-25,-16,37,15,-9,5,33,-9,8,11,-32,-34,-19,-20,14,19,25,19,-36,26,-26,-27,-32,36,-6,-46,-10,34,21,-50,-29,2,-28,-34,-18,0,-31,17,-38,18,37,48,-8,-5,-3,-24,-37,-6,-35,-36,22,-33,-2,-24,0,-33,-19,-30,-22,-10,31,38,29,37,-46,-9,31,46,15,-47,-29,33)  n(34)  
$INVAR a[](5,-24,7,-6,-15,21,18,50,11,-2,-24,-25,47,37,48,-19,-15,-46,3,-3,38,48,48,-16,-4,-7,-44,3,33,-14,25,16,-45,13,38,3,32,18,3,7,-46,39,22,42,23,8,-23,17,27,41,-29,-16,-2,-45,42,-9,46,15,33,1,-16,-34,7,26,48,19,19,-39,12,0,-10,42,-3,-36,-39,35,15,-42,-7,-39,-29,-38,-40,-4,-42,45,-42,-7,-19,33,3,38,5,-37,9,21,28,-39,-34,-12)  n(1068)  
$INVAR a[](-8,48,2,46,-15,-45,-49,33,33,6,29,45,-1,26,35,-23,-40,-47,-23,43,32,-15,33,-17,-11,39,-11,-22,5,-6,47,37,43,-27,-36,36,-49,-10,-8,-50,-35,-33,14,1,5,9,3,-13,30,-16,-37,5,-13,35,-6,-37,-43,41,22,38,3,-13,-26,34,45,26,-4,50,50,29,24,-44,-9,-42,-27,36,18,21,21,44,39,-27,-50,-4,14,-7,30,-13,14,39,22,-16,-23,-47,41,-30,-19,3,26,27)  n(1219)  
$INVAR a[](25,23,-5,0,-26,-46,-40,31,33,46,-14,-50,3,24,-49,1,-8,-8,14,26,-1,24,-20,29,-33,50,-36,-40,1,0,11,13,-6,-29,-29,5,18,43,-37,-28,-27,42,-36,22,-4,46,1,-1,-30,-25,-45,-37,-2,-5,-22,5,21,50,-4,22,-8,-11,48,43,47,-37,-17,-25,33,-8,19,8,-32,33,21,-1,-33,16,-26,26,35,11,45,-25,32,38,49,17,-48,49,-6,-34,-36,42,49,45,22,-25,47,-16)  n(1386)  
$INVAR a[](20,-31,2,6,23,6,-49,45,38,27,-39,46,-50,36,-25,46,-40,-46,33,36,2,39,3,6,-9,6,-3,33,-2,-34,7,22,-48,-15,-40,23,-46,-1,0,-19,49,-22,30,0,-47,-1,-19,26,-5,-29,-9,-30,-35,42,0,50,-4,-50,2,-11,39,31,21,26,-21,-26,-20,26,32,-11,-7,-37,-13,40,4,-34,12,37,32,-1,28,20,23,45,7,49,-1,16,-23,-10,-37,-3,11,-12,-4,-2,42,27,-48,-35)  n(1944)  
$INVAR a[](-42,28,38,26,-29,38,13,-14,31,-49,-47,15,-12,22,-19,-40,41,8,45,42,-43,-46,-43,44,-4,32,8,46,30,-17,6,-46,-1,-44,-31,-49,-28,30,-19,38,-16,-34,-48,-46,-27,34,-34,17,-49,-10,19,-22,-40,30,-49,21,13,14,-47,26,18,-43,-20,-7,38,32,11,-3,21,1,47,-13,16,23,15,3,-37,-44,46,34,12,-31,17,-50,4,41,-6,-3,40,-6,13,10,8,17,-16,10,34,36,7,-29)  n(1054)  
$INVAR a[](47,26,-25,48,22,4,-34,-17,25,14,-21,40,14,-1,29,-19,-44,12,-38,39,-30,-37,0,2,45,29,-50,-31,5,2,-10,1,38,32,16,-30,-3,-26,30,28,50,-50,-2,2,49,-2,9,12,14,38,10,-34,-17,18,0,-35,-37,-13,21,36,-33,22,-48,-22,50,5,24,-22,-38,21,-20,35,-14,40,-18,-1,18,-23,20,25,25,-30,-39,-19,-16,-2,-20,-34,-33,45,32,38,-39,-23,-39,35,1,-19,-31,-33)  n(658)  
$INVAR a[](18,-4,42,-48,-19,46,2,-3,48,46,-12,-2,-17,39,-6,7,-44,-35,7,19,-4,29,47,-18,-43,16,-49,43,5,-40,9,17,13,15,-33,-28,31,48,-21,-31,-20,11,-39,-40,-18,24,-32,-9,20,-34,12,28,-11,2,-8,1,-2,43,-2,-47,13,43,3,25,-39,-7,-27,11,43,24,47,-16,0,-34,-37,-40,-31,17,-50,47,-12,8,-6,6,46,-26,47,30,6,-23,-43,-2,21,19,24,-38,41,-39,-1,-2)  n(215)  
$INVAR a[](-28,-31,-46,24,8,-34,8,-43,29,13,-1,-28,49,21,4,-43,5,20,-27,29,-22,-18,37,26,14,-28,36,-47,-12,-33,33,-14,-45,28,-45,-38,-33,14,-26,32,-32,-27,25,-50,-9,44,24,19,16,48,39,14,-45,-19,15,-31,0,12,-40,-18,21,36,46,-20,7,-22,-25,-46,7,24,33,43,-16,0,10,42,8,-31,-40,-45,49,6,-49,20,39,40,4,-38,26,7,33,-15,21,-49,39,-20,24,45,8,15)  n(755)  
$INVAR a[](-37,-41,48,-35,-32,29,-27,0,-18,50,9,11,-40,32,32,25,45,13,-7,-28,26,-49,-37,-6,-4,11,10,17,-46,18,-40,-17,-23,-8,31,-31,5,-6,10,3,-11,17,24,-11,-6,42,35,-8,-34,16,44,33,-3,26,-11,-1,-47,-17,-19,-9,-48,40,5,31,-8,-7,-31,16,-21,-33,7,0,9,-45,42,23,-12,-22,-30,35,11,-6,48,-28,14,50,45,-13,-8,-32,33,42,41,22,26,12,42,15,-41,-32)  n(1807)  
$INVAR a[](16,16,42,-4,-4,-37,5,-19,-11,31,-49,28,44,-41,-29,39,18,-48,-16,33,-20,5,28,-28,26,49,-9,47,33,-40,-21,-46,-19,13,-11,6,6,-21,38,42,-35,10,-21,15,32,46,-11,14,49,47,15,-2,15,34,14,-13,-32,-19,28,29,1,10,-31,16,19,-41,13,49,34,-47,24,-38,43,-23,-38,-48,34,21,49,-33,33,-15,-32,-28,-30,31,-32,-16,38,18,-34,32,-49,-13,-38,5,-28,-47,47,-18)  n(1195)  
$INVAR a[](-31,-36,-25,30,-17,49,28,-1,12,-26,0,11,36,29,-26,-50,-32,-45,3,-30,-39,43,13,4,-28,9,-23,-38,-50,26,-50,20,44,22,-37,2,48,-25,4,33,24,-10,-46,6,-33,-2,46,-8,8,50,41,-11,-24,-32,-14,-27,49,5,-28,25,-26,-12,-13,26,44,-25,-30,49,42,-44,0,-8,7,39,-47,30,-21,45,44,-23,20,-18,7,-7,22,-24,8,-27,-41,-33,-25,-1,34,18,-10,46,15,-12,-7,-24)  n(837)  
$INVAR a[](-17,-17,-3,8,49,-47,-31,-44,48,22,-36,22,-34,7,-14,-32,39,-11,-4,-38,-40,-7,-12,-5,-48,-29,-32,8,-25,-2,11,32,-32,32,14,11,-38,-8,-12,-25,-17,-31,33,-21,-38,-43,-5,33,-2,-32,16,-24,-50,-4,0,-22,-37,16,-27,-32,34,28,16,28,-17,16,-31,-7,3,-47,-31,-45,-32,-25,-42,-41,48,17,-39,31,-45,-28,39,-16,21,34,50,-2,-28,6,-2,-46,-39,12,25,-22,-33,33,-21,26)  n(486)  
$INVAR a[](-11,22,-21,6,33,-23,3,-38,41,-7,47,-48,-30,-29,27,47,-5,30,19,-40,-22,40,-37,-32,11,-21,6,-39,23,-41,32,-22,-12,29,-21,-30,-1,-19,-31,-7,-50,8,-25,23,-1,-19,4,-47,-47,48,-11,12,20,-26,3,-4,29,-11,38,37,-11,-21,13,24,-40,-39,16,-46,38,41,-26,-9,-13,2,25,41,-13,13,19,-10,41,40,-15,-46,-15,-10,47,41,12,39,36,9,-42,4,-11,-2,-49,-15,0,-49)  n(1071)  
$INVAR a[](9,9,-18,41,-13,-14,15,-5,3,-12,-49,-15,-22,-40,2,-8,19,41,36,37,14,48,21,-31,19,33,-50,49,-14,48,6,43,23,-49,-38,17,-22,-40,48,-4,-11,4,-5,7,41,44,-35,42,24,-25,19,-4,-11,17,-36,-15,28,43,21,21,40,34,-35,11,-36,45,-49,30,-25,4,24,50,-27,8,-38,41,37,-30,29,-24,-3,32,13,10,-46,50,-22,0,-43,24,-30,1,5,19,40,-40,49,-4,-19,-36)  n(1008)  
$INVAR a[](44,-23,36,43,-38,-27,35,7,48,-47,-48,6,34,-40,26,-40,-16,-33,-17,-43,-31,49,-12,10,40,6,-27,6,-3,-1,-29,19,17,23,20,35,-19,48,7,-1,22,-17,-33,16,-21,-25,-46,29,37,6,48,-24,-23,-19,-26,15,-29,19,42,20,5,-35,31,31,-10,18,-26,7,31,14,33,-12,28,32,44,-10,47,-5,-3,-19,-13,-5,-23,9,23,-44,-49,-17,6,23,24,34,49,32,-37,-8,24,0,-25,-28)  n(212)


Seveda je disk space spet na 0b. :D

njok ::

Našel vzrok.

Datoteka C:\WINDOWS\System32\LogFiles\trace.log naraste na ~ 5GB.

A ma tukaj critticall kaj vmes al je to moj problem?

Thomas ::

To bomo pogledali, za vsak slučaj. Ne bi smelo.

Jst ::

Po 15ih urah se mi je zgeneriralo naslednje:


// The algorithm has been enhanced For 59.7884%

 $DECLAREINT n one i j acritticall1 critticall2 critticall4 

 $DIMENSION a[100]
 $INVAR n(100) a[](47,-30,2,-37,38,-48,-2,46,-18,27,19,38,-20,12,33,17,-3,47,-28,46,5,48,-37,-5,50,-31,19,17,-13,-43,32,-8,-4,30,7,-4,-41,45,14,40,-37,29,48,-5,-33,-26,40,3,34,36,41,23,27,-27,-24,-21,26,31,-45,1,-10,-46,-40,-25,-5,-31,-28,-47,17,-19,-8,-30,30,-20,-25,-13,-46,48,-19,-46,-41,-2,6,48,18,9,-46,-50,-38,25,-31,-46,-14,28,16,21,-9,20,-2,24)
 $INVAR n(100) a[](-9,-42,-41,-49,-8,29,-27,-35,29,-37,49,15,37,32,-1,-31,-38,-43,43,22,-39,39,39,-36,40,-14,-19,48,12,-2,-1,-34,16,33,30,43,-12,-47,-22,-1,-19,-41,-28,26,4,19,-33,6,-8,-24,-26,-50,50,-6,8,36,17,47,-17,-33,-42,1,3,10,-24,17,6,-50,-16,-32,-9,-21,3,-18,23,22,-23,25,-29,49,-5,-16,-28,41,-21,45,-16,-6,-15,36,-14,-12,10,11,-48,-44,10,-18,-19,26)
 $INVAR n(100) a[](-30,17,-23,18,-12,-14,7,35,49,-5,-16,-44,9,-46,49,40,31,-23,40,-14,14,-14,-7,15,21,-10,5,-26,3,-44,38,-37,-34,-21,-36,46,32,34,-27,-10,-14,-40,-4,-40,21,21,6,43,10,12,10,-22,-26,39,45,45,24,-14,30,5,-2,29,47,6,16,-35,46,15,30,-6,10,9,35,35,9,-40,-27,-26,34,-13,-50,1,-5,-42,38,43,25,-27,27,29,13,-42,-16,41,-34,40,-9,20,-22,-2)
 $INVAR n(100) a[](-21,46,41,10,16,14,40,29,-43,-38,11,-42,13,30,-42,33,-23,-47,-49,-10,15,30,-49,16,5,-4,47,-3,-12,18,-17,-14,-18,-27,33,4,30,-25,19,5,17,-9,7,-14,-19,17,29,29,-30,37,-14,-21,-50,-17,-41,30,-45,-45,25,-45,34,-29,46,-7,38,-30,-13,-2,49,-18,-10,1,48,26,-22,46,-32,12,-1,-5,-8,-37,26,20,-31,-32,-37,21,-19,36,29,37,-23,21,14,-49,50,38,50,-4)
 $INVAR n(100) a[](25,-23,15,0,-4,-14,-28,-43,15,0,23,0,23,15,9,-10,-12,46,5,-20,-20,13,-26,10,-3,42,3,15,21,-15,1,-4,-13,-12,-36,-48,-39,-47,24,-14,-14,-3,23,31,41,-22,13,-42,-3,-42,37,-28,1,21,-33,-13,-4,43,-13,-43,-41,30,-37,-10,39,25,26,-18,-8,-35,-42,44,-26,-48,-19,-2,29,50,-49,-9,-35,33,31,29,46,50,0,29,6,-37,11,-28,-30,-6,-4,-33,-39,-20,0,-19)
 $RINVAR a[](-50,50) n(1,2000) 
 $RETVAR a[] 
 $WEIGHTS lines=1 commands=1

// int a[100]; int n=0;int one=0;int i=0;int j=0;int acritticall1=0;int critticall2=0;int critticall4=0;

acritticall1=100;
if (n>acritticall1) {
    n=100;
}
$BES
critticall2=75;
critticall2+=-8;
acritticall1=1;
j=a[critticall2];
one=a[critticall4];
i=critticall2;
if (one>j) {
    i+=2;
    critticall2+=-8;
} else {
    critticall2+=-1;
    one+=2;
    j=a[critticall2];
    if (one>j) {
        critticall2+=-18;
    }
}
if (critticall4>j) {
    i=critticall2;
}
labelcritticall15:;
while (i<n) {
    critticall4=i-critticall2;
    one=i;
    i+=1;
    j=a[one];
    critticall4=a[critticall4];
    if (critticall4<=j) {
        goto labelcritticall15;
    }
    a[one]=critticall4;
    one=one-critticall2;
    a[one]=j;
}
critticall2=n;
while (critticall2!=acritticall1) {
    critticall2=sqrt(critticall2);
    i=critticall2;
    while (i<n) {
        critticall4=i-critticall2;
        j=a[i];
        one=i;
        critticall4=a[critticall4];
        i+=1;
        while (critticall4>j) {
            a[one]=critticall4;
            one=one-critticall2;
            if (one<critticall2) {
                break;
            }
            critticall4=one-critticall2;
            critticall4=a[critticall4];
        }
        a[one]=j;
    }
}
$EES
//$RINVAR  a[](-50,50) n(1,2000) 
$INVAR a[](-21,-40,-8,-42,41,-44,5,-1,-27,-42,45,3,-50,6,-5,18,-34,-40,-50,-7,-3,43,33,18,36,11,33,38,21,46,-20,7,-46,22,-43,-4,-35,-7,28,-47,-21,-2,-44,-5,-18,-25,8,-34,15,-35,-14,10,8,-3,38,24,-39,14,14,-5,30,-8,1,-46,46,35,32,-29,28,-49,-47,-14,26,-26,10,-29,5,47,-36,37,-12,-10,-27,44,-11,-39,-16,-23,17,11,17,7,25,39,-48,33,47,6,46,-32)  n(1744)  
$INVAR a[](-18,18,23,44,4,-21,35,40,6,-35,-28,48,5,-22,-6,-20,10,0,-20,-19,-28,-2,-5,-43,-6,-1,50,-40,15,-12,48,24,-41,-36,45,-39,8,-45,-24,-21,22,-31,16,-38,40,50,-39,32,26,38,-7,-21,44,-16,20,44,-36,49,21,-32,-42,-19,-49,-44,11,-21,-40,24,-23,-45,47,0,-50,-26,46,46,14,24,35,-25,19,48,-13,-3,-30,47,7,28,-47,18,46,35,9,6,-23,-1,44,-29,-46,30)  n(1769)  
$INVAR a[](-29,5,-13,-18,-12,-47,0,31,-49,-34,30,3,-35,42,10,43,46,41,32,43,-21,-9,47,-24,3,48,19,13,-25,-11,-5,13,-2,-19,-19,46,8,43,28,-28,22,-27,-30,-14,42,-44,-20,-1,29,31,7,25,2,-11,37,-13,41,-38,-10,-40,-22,-11,21,28,12,-21,18,-8,-8,18,-9,-3,41,-19,12,41,18,47,43,-31,46,24,50,9,-47,41,20,-40,-4,-38,0,-37,-39,21,31,9,-8,-15,-35,-46)  n(1)  
$INVAR a[](-26,25,-46,-46,-21,19,-18,48,7,-11,-32,-45,26,5,-38,-16,36,32,-50,9,5,44,49,-45,-40,-38,-32,40,-5,50,4,49,45,-22,37,-50,41,44,43,-40,42,10,42,18,-3,-18,46,36,4,-30,-18,33,-48,-23,-5,-35,-25,11,-43,5,42,36,49,44,-35,-17,-25,18,26,-1,40,-25,-1,6,-8,43,-32,-11,10,47,-45,46,-7,12,-16,-39,-23,43,-19,-9,-8,41,-33,-37,28,34,-11,43,-28,15)  n(1181)  
$INVAR a[](22,-28,-11,29,-3,-25,25,-47,44,10,-8,-25,-32,-21,-4,0,0,30,13,27,9,-11,-31,-12,43,-9,25,15,-34,-3,50,-47,-24,-42,1,-3,-44,-20,36,-2,-29,-1,24,-25,30,-39,49,47,48,-30,-8,36,8,-41,-31,35,1,-23,36,31,-35,18,-14,23,-3,21,37,-16,-28,17,37,50,12,20,-40,35,-10,-41,-27,11,40,23,-14,2,34,50,18,-32,42,-9,50,-7,-28,40,34,-4,-33,-29,-11,33)  n(1094)  
$INVAR a[](-11,-23,-6,-46,38,28,-21,-17,-9,9,33,-16,-37,-33,-44,22,45,19,-16,3,40,-47,17,-41,-16,-13,-4,-13,-9,37,-3,19,-14,22,-25,-50,11,35,13,-37,8,22,38,16,31,50,13,-19,24,-12,22,-29,-18,4,-6,-40,-27,-49,-47,6,-21,18,-28,34,13,40,-29,22,-9,16,6,15,-23,-29,6,38,36,-41,-17,8,45,-14,14,-41,-29,18,-44,7,5,-1,-29,49,-5,-32,41,-36,43,-7,21,-44)  n(11)  
$INVAR a[](2,-32,48,-42,3,-23,-31,16,-8,-38,-31,38,25,-46,-4,2,-32,-21,41,3,-16,30,1,29,16,13,-49,11,-40,12,27,31,31,-34,10,-8,17,-5,-2,-23,9,-9,-47,-45,46,-39,-20,17,-10,13,-13,30,-40,-23,26,-35,-44,45,8,32,-42,-41,43,8,-3,-33,-47,-30,25,39,-39,-50,16,5,43,-21,2,-16,7,-11,4,-30,34,-15,16,26,-15,16,-34,0,-22,-4,50,-49,-34,-44,3,-34,-10,36)  n(636)  
$INVAR a[](-4,21,-33,18,-47,13,-6,20,33,42,-43,-38,-4,23,-21,38,-13,37,9,-22,-47,49,49,-7,38,13,-49,-5,-6,22,-18,9,19,38,35,-20,19,9,43,-45,-38,-33,38,-26,-32,-39,27,0,32,-33,9,-4,15,-27,43,3,20,-50,36,-9,36,40,23,-22,41,18,48,48,-33,31,-32,6,0,3,-50,-41,39,-39,21,5,46,28,-12,-46,-14,-25,-34,-45,-4,-43,41,-34,-45,33,-24,-14,-44,-29,-48,-24)  n(538)  
$INVAR a[](32,-41,-40,26,-29,-6,-18,-47,-8,48,-24,42,-3,-12,-9,12,-33,-9,-48,49,-21,-29,20,30,-37,2,34,-4,-1,-23,-10,-40,38,3,-11,30,-30,-45,-27,3,1,28,36,-32,-10,41,-49,11,-30,13,-13,3,-46,5,-18,36,-12,-31,-3,30,21,-3,-21,4,-11,28,9,19,37,-44,-11,17,27,-26,22,6,12,9,-20,-10,-42,-40,-49,21,-39,48,36,24,-18,3,2,45,-6,-7,-47,-4,31,-49,45,4)  n(178)  
$INVAR a[](-50,-27,19,-15,14,11,19,17,9,12,-8,-9,-36,15,-6,13,-45,-5,20,-23,-33,31,-17,-3,-4,15,18,0,-39,-49,5,44,40,23,-49,45,43,10,40,25,-3,8,37,15,8,-37,45,-23,8,34,11,29,22,34,47,-24,0,30,3,13,11,-3,23,9,-24,40,-36,14,1,44,50,-41,37,25,-1,8,-29,-24,49,4,33,28,-36,21,-46,-18,-1,3,47,-14,-46,-50,-46,48,44,2,11,8,-31,0)  n(1138)  
$INVAR a[](40,11,23,20,19,-24,44,-1,-21,-40,-48,2,-46,-37,12,18,44,-32,11,-5,-17,-39,10,7,-29,-19,-14,22,48,-16,15,-38,3,6,-46,-45,-42,45,8,9,-46,-6,-11,-40,25,12,12,-13,-46,2,-23,-27,44,35,35,-29,29,39,40,8,35,-50,22,18,49,-13,-41,0,48,-18,8,47,29,15,19,-27,7,5,-19,-48,-13,-39,34,42,31,3,-48,45,-39,4,7,48,-7,4,31,27,-28,13,-26,20)  n(2)  
$INVAR a[](-2,35,-21,-1,-20,4,4,-19,45,29,24,-47,-26,50,17,50,-21,10,-32,-32,-15,31,7,29,-7,25,20,-16,-8,-20,17,-48,26,8,-12,31,44,16,-29,31,-8,-13,11,46,24,46,8,35,32,47,-50,-13,39,35,2,-3,-10,41,-16,-35,44,23,-35,14,38,-32,-32,-31,-2,5,29,-47,12,3,-12,2,-45,-43,2,-29,-4,-22,-17,31,-28,34,23,35,29,29,-41,31,-33,5,-21,46,42,-9,-22,45)  n(900)  
$INVAR a[](8,12,-43,37,-6,-21,8,36,19,27,3,48,47,-20,-20,6,-11,18,1,35,-18,-33,49,6,-3,45,15,1,-28,-1,-4,44,-34,-1,-7,34,-7,-39,-7,22,-39,-12,-15,-10,-49,-2,17,35,-10,46,17,-38,13,-37,28,50,16,-19,22,26,44,7,3,-27,-30,-43,21,-31,-49,-5,27,-5,-25,-4,-29,2,33,-44,29,9,-28,19,-39,50,35,-36,-5,26,-43,19,-37,19,12,49,19,-42,16,-39,48,-40)  n(1016)  
$INVAR a[](-38,13,31,34,-49,-16,-21,27,-20,-32,-21,-2,40,-11,-22,-35,-11,0,7,4,-33,-42,-37,23,1,-1,-36,3,50,48,16,13,21,-45,-34,38,-39,44,29,-20,9,-13,4,23,43,-10,34,-41,-15,15,47,32,37,-29,18,39,48,-4,16,-10,-47,49,20,34,35,24,47,-24,-28,-42,-1,-34,16,-32,33,28,-27,-40,1,47,10,-14,42,-48,-5,50,-14,-39,41,-23,25,48,30,15,-43,-24,14,5,1,24)  n(1896)  
$INVAR a[](-10,2,-44,43,-5,49,16,5,41,-30,6,-33,-22,-5,26,36,47,-39,16,35,-41,-39,-7,-12,-14,37,32,-5,-28,-35,41,43,-6,43,-36,-32,-48,40,2,-9,-46,-19,22,29,12,-6,16,1,-1,6,12,-14,10,0,-31,-45,19,-47,-15,41,46,16,10,-27,-32,5,50,22,40,35,-1,15,-9,-20,30,9,41,-28,33,-35,41,-41,-10,27,19,-21,16,5,12,41,8,35,-17,-13,-13,50,-19,22,12,-18)  n(771)
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Thomas ::

Link za trace.

Sicer pa ... tele sorte hvaležno zbiram in bom naredil nekaj iz njih ...

Eno ilustracijo ...

njok ::

Trace.log je ze odpravljen. Problem je bil pa program bootvis, ki je stalno pisal v tisto datoteko, ne vem pa, zakaj ga je laufanje critticalla toliko pospesilo.

Se opravicujem, ker sem malo posmetil temo. No ja, mogoce bo pa se kdo mel isti problem. ;)

Jst ::

Thomas: in kaj boš naredil z zbranimi sorti?

..

Za vse, ki ste veckrat laufali sort quest in ne veste, kje v datoteki fs.c_e je najhitrejši sort, sem napisal en programcek, ki vam to najde namesto vas. Nisem se spuščal v to, katere package moram odstraniti (napisano v Delphiju v parih minutah), zato je exe neverjetno velik (400 KB).

FindBest.exe
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Jst ::

Pa tudi če laufate prvič, ni nujno, da je zadnji sort najhitrejši.
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Thomas ::

njok,

No problem whatsoever.

Jst,

Good. Zeitplantag gre h koncu ... do vecera bom naredil ilustracijo. Ko jo boš videl, boš razumel. ;)

Sicer pa ... cenim vse kar počneš. Se strinjam, s tvojimi zadnjimi posti tud.

Thomas ::

Tukaj sta dva paralelna diagrama. Sort, ki sem ga dal kot inicialni sort je na levi, na desni pa Romcijeva izbolšana verija.

Na začetku, v prvi vrstici slike so dvakrat razmetani elementi arraya.

Prvi $invar a[] ...

Druga vrstica je drugi korak programa. Osnovnega (fs.c) na levi in Romcijevega na desni.

Bolj greste proti dnu slike, bolj so sivi toni urejeni, poferlaufani.

Vsaka linija nižje, kvečjemu bolj urejen array a[], en korak v programu naprej.

Zelene pike pomenijo branje iz arraya a[], rdeče pisanje vanj. Iz/na tisti index, kjer pika je.

A štekate? A je rox a je sux?

Zgodovina sprememb…

  • zavarovalo slike: OwcA ()

Vesoljc ::

prelep 3d film
Abnormal behavior of abnormal brain makes me normal...

njok ::

Glede na to, da je slika velika 1,2MB bi bilo lepo, ce bi se podalo do nje samo link.

Drugace pa zelo zanimivo, zgleda kot da je pogruntal, da lahko tisti del, ko na koncu parkrat prečese celoten array, preleti samo dvakrat. Pa proti koncu skace po arrayu priblizno po sinusoidi. Vsekakor zelo lep graf!

Zgodovina sprememb…

  • spremenil: njok ()

Thomas ::

To smo zdej dodal v Critticall. $image ukaz, ki ponazori, kaj se je dogajalo z arrayem v programu.

Telih sortov je toliko, kolikor je gnujev na afriški savani. Jih bomo vse poslikal, pa še nove izvabili iz lukenj, kjer se še skrivajo.

Sem pa dal zdej manjšo sliko gor. Vsak jo lahko raztegne potem sam, za faktor 2 ali tri po x in y osi.

Jst ::

Katerega arraya? A ni bilo na začetku 5 arrayev?

In ta, romčijev sort, velja samo za porazdelitev -50,50?
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|

Thomas ::

Arraya, ki ga izbereš. Ta se poslika. Katerikoli od 16.

Velja tudi za -50 do 50. Ni kakšne razlike od -10 do 10.

Slikamo ...

Jst ::

Missed my point.

Mene zanima, če ta romčijev sort da pravilne rezultate na katerikoli porazdelitvi?

btw: kako napreduje nova verzija critticalla?
(moj CPU je spet osamljen)
Islam is not about "I'm right, you're wrong," but "I'm right, you're dead!"
-Wole Soyinka, Literature Nobelist
|-|-|-|-|Proton decay is a tax on existence.|-|-|-|-|
««
22 / 29
»»

Vredno ogleda ...

TemaSporočilaOglediZadnje sporočilo
TemaSporočilaOglediZadnje sporočilo
»

Najhitrejši programski jezik? (strani: 1 2 )

Oddelek: Programiranje		757755 (5575) Senitel
»

Funkcija z logičnimi operaterji.... (strani: 1 2 )

Oddelek: Programiranje		905571 (4917) CaqKa
»

Petaflopsu naproti (strani: 1 2 3 )

Oddelek: Novice / Procesorji		1058877 (8877) Marjan
»

cene permutacij help please

Oddelek: Programiranje		262076 (1683) Sergio
»

kako definirtati prastevilo

Oddelek: Programiranje		143799 (3604) ooux

Več podobnih tem