<?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>sujet From Ipp32f* to double** dans Intel® Integrated Performance Primitives</title>
    <link>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825971#M5176</link>
    <description>Hi there,&lt;BR /&gt;&lt;BR /&gt;I need some help here, is it possible to go from a Ipp32f* to a double**??? I'm trying to plot a 3D surface using QwtPlot3D and the loadFromData function ask my data to be like a double pointer. In other words, they are expecting a matrix and not a vector like I have my Ipp data. Let's say that my IPP vector has 20 elements (1x20) and the loadFormData function expects something like data[5][4].&lt;BR /&gt;&lt;BR /&gt;So my question is, is there any way to transform an IPP vector into a double** without creating a loop within a loop to assign element by element the values of the matrix???&lt;BR /&gt;&lt;BR /&gt;Thanks for your help,&lt;BR /&gt;&lt;BR /&gt;Boris</description>
    <pubDate>Fri, 10 Sep 2010 11:39:20 GMT</pubDate>
    <dc:creator>apolo74</dc:creator>
    <dc:date>2010-09-10T11:39:20Z</dc:date>
    <item>
      <title>From Ipp32f* to double**</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825971#M5176</link>
      <description>Hi there,&lt;BR /&gt;&lt;BR /&gt;I need some help here, is it possible to go from a Ipp32f* to a double**??? I'm trying to plot a 3D surface using QwtPlot3D and the loadFromData function ask my data to be like a double pointer. In other words, they are expecting a matrix and not a vector like I have my Ipp data. Let's say that my IPP vector has 20 elements (1x20) and the loadFormData function expects something like data[5][4].&lt;BR /&gt;&lt;BR /&gt;So my question is, is there any way to transform an IPP vector into a double** without creating a loop within a loop to assign element by element the values of the matrix???&lt;BR /&gt;&lt;BR /&gt;Thanks for your help,&lt;BR /&gt;&lt;BR /&gt;Boris</description>
      <pubDate>Fri, 10 Sep 2010 11:39:20 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825971#M5176</guid>
      <dc:creator>apolo74</dc:creator>
      <dc:date>2010-09-10T11:39:20Z</dc:date>
    </item>
    <item>
      <title>From Ipp32f* to double**</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825972#M5177</link>
      <description>Hi Boris,&lt;BR /&gt;&lt;BR /&gt;I think, that you will have to go through two loops. The reason is, that you need to convert an array of 32-bit data (Ipp32f) into 64-bit data (double); first loop. Then you can assign the individual pointers in your array of double** to map the correspondings vectors in your converted 1D-array; second short loop.&lt;BR /&gt;&lt;BR /&gt;Best regards, TJ&lt;BR /&gt;</description>
      <pubDate>Fri, 10 Sep 2010 13:29:55 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825972#M5177</guid>
      <dc:creator>Thomas_B_3</dc:creator>
      <dc:date>2010-09-10T13:29:55Z</dc:date>
    </item>
    <item>
      <title>From Ipp32f* to double**</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825973#M5178</link>
      <description>Hi TJ,&lt;BR /&gt;&lt;BR /&gt;thanks for the reply... let's say that I work with Ipp64f* instead of Ipp32f* to avoid the transformation between float and double, do I still need to assign one by one the values between my vector and the double pointer? I just wanted to avoid loops and allocation and deallocation of the double pointer.&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;Ipp64f xx[5*4] = { 10, 0, 5, 0, 10, &lt;BR /&gt;        0, 0, 4, 0, 0, &lt;BR /&gt;        0, 0, 3, 0, 0, &lt;BR /&gt;        10, 0, 2, 0, 10}; &lt;BR /&gt;double **x; &lt;BR /&gt;x = new double*[4]; &lt;BR /&gt;for(int i=0; i&amp;lt;5; i++)&lt;BR /&gt; x&lt;I&gt; = new double[4]; &lt;BR /&gt;&lt;BR /&gt;// --- these are the loops I'd love to avoid ---&lt;BR /&gt;for(int i=0; i&amp;lt;4; i++)&lt;BR /&gt; for(int j=0; j&amp;lt;5; j++)&lt;BR /&gt;  x&lt;J&gt;&lt;I&gt; = xx[5*i+j]; &lt;BR /&gt;&lt;BR /&gt; .... [do something with the double pointer] ...&lt;BR /&gt;&lt;BR /&gt;for( int i=0; i&amp;lt;5; i++) &lt;BR /&gt; delete x&lt;I&gt;; &lt;BR /&gt;delete []x; &lt;BR /&gt;&lt;BR /&gt;Boris&lt;BR /&gt;&lt;/I&gt;&lt;/I&gt;&lt;/J&gt;&lt;/I&gt;</description>
      <pubDate>Fri, 10 Sep 2010 13:58:46 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825973#M5178</guid>
      <dc:creator>apolo74</dc:creator>
      <dc:date>2010-09-10T13:58:46Z</dc:date>
    </item>
    <item>
      <title>From Ipp32f* to double**</title>
      <link>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825974#M5179</link>
      <description>Hi Boris,&lt;BR /&gt;&lt;BR /&gt;I was thinking of something like this (I didn't test it):&lt;BR /&gt;&lt;BR /&gt;Ipp64f xx[5*4] = { 10, 0, 5, 0, 10, &lt;BR /&gt;        0, 0, 4, 0, 0, &lt;BR /&gt;        0, 0, 3, 0, 0, &lt;BR /&gt;        10, 0, 2, 0, 10}; &lt;BR /&gt;double **x; &lt;BR /&gt;x = new double*[4]; &lt;BR /&gt;&lt;BR /&gt;// --- these are the loops I'd love to avoid ---&lt;BR /&gt;for(int i=0; i&amp;lt;4; i++)&lt;BR /&gt;  x&lt;I&gt; = &amp;amp;xx[5*i];&lt;BR /&gt;&lt;BR /&gt; .... [do something with the double pointer] ...&lt;BR /&gt;&lt;BR /&gt;delete []x;&lt;BR /&gt;&lt;BR /&gt;With this approach you avoid copying your data. **x is pointing to your "original" data but by accessing via x it &lt;I&gt;appears &lt;/I&gt;to be reorganised as a matrix.&lt;BR /&gt;&lt;BR /&gt;&lt;BR /&gt;Best regards,&lt;BR /&gt;TJ&lt;/I&gt;</description>
      <pubDate>Fri, 10 Sep 2010 15:44:57 GMT</pubDate>
      <guid>https://community.intel.com/t5/Intel-Integrated-Performance/From-Ipp32f-to-double/m-p/825974#M5179</guid>
      <dc:creator>Thomas_B_3</dc:creator>
      <dc:date>2010-09-10T15:44:57Z</dc:date>
    </item>
  </channel>
</rss>

