<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:taxo="http://purl.org/rss/1.0/modules/taxonomy/" version="2.0">
  <channel>
    <title>topic Re: mkl fmgres :  try to solve 1 euquation with 1 unknown in Intel® oneAPI Math Kernel Library</title>
    <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-fmgres-try-to-solve-1-euquation-with-1-unknown/m-p/907450#M11851</link>
    <description>Hi, reply to myself for those who try the same,&lt;BR /&gt;&lt;BR /&gt;1. I found that the declaration of the tmp array: tmp[N*(2N+1)+(N*(N+9))/2+1] is less general then it looks. For N=1 this generates a too small tmp array, since the code actually uses a longer array. Thus for small N do not use that declaration, but add something extra, eg tmp[100+N*(2N+1)+(N*(N+9))/2+1].&lt;BR /&gt;&lt;BR /&gt;2. Also notice that RCI_request==3 contains code dependend on the size N.&lt;BR /&gt;</description>
    <pubDate>Fri, 28 Sep 2007 10:20:20 GMT</pubDate>
    <dc:creator>dennisgerardusroddem</dc:creator>
    <dc:date>2007-09-28T10:20:20Z</dc:date>
    <item>
      <title>mkl fmgres :  try to solve 1 euquation with 1 unknown</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-fmgres-try-to-solve-1-euquation-with-1-unknown/m-p/907449#M11850</link>
      <description>Hi, beginner problem, I try to solve A x = b with A = 0.2 and b=0.1 so x should become 0.5. In the fgmres_full_funct_c.c example I did this:&lt;BR /&gt;&lt;BR /&gt;#define N 1&lt;BR /&gt;int ia[2] = {1, 2};&lt;BR /&gt;int ja[1] = { 1 };&lt;BR /&gt;double A[1] = { 0.2 };&lt;BR /&gt;double expected_solution&lt;N&gt; = { 0.5 };&lt;BR /&gt;&lt;BR /&gt;when I compile and run the fgmres_full_funct_c.out I get a 'nan' in the computed solution in stead of 0.5 I use mkl version 9.0.018. &lt;BR /&gt;&lt;BR /&gt;thanks, Dennis&lt;BR /&gt;&lt;/N&gt;</description>
      <pubDate>Mon, 24 Sep 2007 07:57:25 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-fmgres-try-to-solve-1-euquation-with-1-unknown/m-p/907449#M11850</guid>
      <dc:creator>dennis_r_</dc:creator>
      <dc:date>2007-09-24T07:57:25Z</dc:date>
    </item>
    <item>
      <title>Re: mkl fmgres :  try to solve 1 euquation with 1 unknown</title>
      <link>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-fmgres-try-to-solve-1-euquation-with-1-unknown/m-p/907450#M11851</link>
      <description>Hi, reply to myself for those who try the same,&lt;BR /&gt;&lt;BR /&gt;1. I found that the declaration of the tmp array: tmp[N*(2N+1)+(N*(N+9))/2+1] is less general then it looks. For N=1 this generates a too small tmp array, since the code actually uses a longer array. Thus for small N do not use that declaration, but add something extra, eg tmp[100+N*(2N+1)+(N*(N+9))/2+1].&lt;BR /&gt;&lt;BR /&gt;2. Also notice that RCI_request==3 contains code dependend on the size N.&lt;BR /&gt;</description>
      <pubDate>Fri, 28 Sep 2007 10:20:20 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/mkl-fmgres-try-to-solve-1-euquation-with-1-unknown/m-p/907450#M11851</guid>
      <dc:creator>dennisgerardusroddem</dc:creator>
      <dc:date>2007-09-28T10:20:20Z</dc:date>
    </item>
  </channel>
</rss>

