Path: csiph.com!usenet.pasdenom.info!weretis.net!feeder4.news.weretis.net!rt.uk.eu.org!newsfeed.xs4all.nl!newsfeed4.news.xs4all.nl!xs4all!post.news.xs4all.nl!not-for-mail
To: python-list@python.org
From: Joseph Martinot-Lagarde <joseph.martinot-lagarde@m4x.org>
Subject: Re: How can this assert() ever trigger?
Date: Mon, 12 May 2014 19:05:25 +0200
References: <536e44c1$0$27147$e4fe514c@dreader35.news.xs4all.nl>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:24.0) Gecko/20100101 Thunderbird/24.5.0
In-Reply-To: <536e44c1$0$27147$e4fe514c@dreader35.news.xs4all.nl>
Precedence: list
Newsgroups: comp.lang.python
Message-ID: <mailman.9917.1399914607.18130.python-list@python.org>
Lines: 63
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: csiph.com comp.lang.python:71395

Le 10/05/2014 17:24, Albert van der Horst a écrit :
> I have the following code for calculating the determinant of
> a matrix. It works inasfar that it gives the same result as an
> octave program on a same matrix.
>
> / ----------------------------------------------------------------
>
> def determinant( mat ):
>      ''' Return the determinant of the n by n matrix mat
>          i row j column
>          Destroys mat ! '''
>      #print "getting determinat of", mat
>      n=len(mat)
>      nom = 1.
>      if n == 1: return mat[0][0]
>      lastr = mat.pop()
>      jx=-1
>      for j in xrange(n):
>         if lastr[j]:
>             jx=j
>             break
>      if jx==-1: return 0.
>      result = lastr[jx]
>      assert(result<>0.)
>      # Make column jx zero by subtracting a multiple of the last row.
>      for i in xrange(n-1):
>          pivot = mat[i][jx]
>          if 0. == pivot: continue
>          assert(result<>0.)
>          nom *= result   # Compenstate for multiplying a row.
>          for j in xrange(n):
>              mat[i][j] *= result
>          for j in xrange(n):
>              mat[i][j] -= pivot*lastr[j]
>      # Remove colunm jx
>      for i in xrange(n-1):
>         x= mat[i].pop(jx)
>         assert( x==0 )
>
>      if (n-1+jx)%2<>0: result = -result
>      det = determinant( mat )
>      assert(nom<>0.)
>      return result*det/nom
>
> /-----------------------------------------
>
> Now on some matrices the assert triggers, meaning that nom is zero.
> How can that ever happen? mon start out as 1. and gets multiplied
> with a number that is asserted to be not zero.
>
> Any hints appreciated.
>
> Groetjes Albert
>
I know it's not the question, but if you want a replacement for octave 
did you try numpy (and scipy) ? The determinant would be computer faster 
and with less memory than with your function.

---
Ce courrier électronique ne contient aucun virus ou logiciel malveillant parce que la protection avast! Antivirus est active.
http://www.avast.com