<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
  <head>
    <meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#ffffff" text="#000000">
    Nick and Daniel,<br>
    <br>
    Thanks for the suggestions. I did manage to get it to run after
    playing with some of the variables. If I used the default values for
    "sample" and "min_size" it worked. Unfortunately the default values
    are less than ideal. I think the main problem is with the "sample"
    parameter. When I enter "sample=5,5" I assume that every fifth pixel
    from every fifth line is being sampled for use in calculating the
    cluster statistics. This is apparently too many samples for
    i.cluster to handle. If I use "sample=100,100" the program works as
    expected (more or less). I have a feeling that the sample array is
    not able to handle a large number of samples. <br>
    <br>
    Thanks again,<br>
    <br>
    Ned<br>
    <br>
    <br>
    On 02/18/2011 09:22 AM, Nick Jachowski wrote:
    <blockquote
      cite="mid:B037D1F1-F0CD-4489-8EC6-4BD3C939164B@gmail.com"
      type="cite">
      <div>Also, make sure you mask any Nulls since that often seems to
        be the problem when i.cluster doesn't work as expected,
        especially if you are using landsat 7 slc-off imagery - although
        I've also experienced it with Landsat 4 and 5 imagery. <br>
        <br>
        [Sent from a mobile device]</div>
      <div><br>
        On Feb 18, 2011, at 4:38 PM, Nick Jachowski &lt;<a
          moz-do-not-send="true" href="mailto:njachowski@gmail.com">njachowski@gmail.com</a>&gt;
        wrote:<br>
        <br>
      </div>
      <blockquote type="cite">
        <div>Hi Ned and Daniel,
          <div><br>
          </div>
          <div>I have run into this problem in the past, and I believe
            it may have to do with your convergence, min_size, and
            classes settings.  Try decreasing some of those and see if
            you don't get better results.  You can also try increasing
            your separation from 0.0 (the default) to 1.0 or 1.5.</div>
          <div><br>
          </div>
          <div>Nick</div>
          <div>
            <div><br>
              <div class="gmail_quote">On Fri, Feb 18, 2011 at 4:24 PM,
                daniel mcinerney <span dir="ltr">&lt;<a
                    moz-do-not-send="true"
                    href="mailto:daniel.o.mcinerney@gmail.com">daniel.o.mcinerney@gmail.com</a>&gt;</span>
                wrote:<br>
                <blockquote class="gmail_quote" style="margin: 0pt 0pt
                  0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204);
                  padding-left: 1ex;">Hi Ned,<br>
                  <br>
                  I have been using i.cluster in GRASS 7 on a 64-bit
                  fedora server,<br>
                  without any trouble.  However, I managed to replicate
                  your problem<br>
                  and create a single class output by setting the
                  parameter<br>
                  'sample=5,5' as you had done in your example. If I run
                  i.cluster without<br>
                  that parameter, the signature file and classification
                  output both <br>
                  have the number of predefined classes. <br>
                  <br>
                  Can you see if this helps with your data?<br>
                  <br>
                  Regards,<br>
                  <font color="#888888">Daniel.</font>
                  <div>
                    <div>
                    </div>
                    <div class="h5"><br>
                      <br>
                      <div class="gmail_quote">On Thu, Feb 17, 2011 at
                        7:49 PM, Ned Horning <span dir="ltr">&lt;<a
                            moz-do-not-send="true"
                            href="mailto:horning@amnh.org">horning@amnh.org</a>&gt;</span>
                        wrote:<br>
                        <blockquote class="gmail_quote" style="margin:
                          0pt 0pt 0pt 0.8ex; border-left: 1px solid
                          rgb(204, 204, 204); padding-left: 1ex;">Hi - I
                          would like to do an unsupervised
                          classification using i.cluster from GRASS 7 on
                          my Ubuntu 64-bit machine. When I run i.cluster
                          then i.maxlik on a Landsat image I only get
                          one output class, the entire image. Only one
                          class appears in the signature file so the
                          problem is with the i.cluster step. I tried
                          running i.cluster on different images and
                          always get just one class.<br>
                          <br>
                          Here is the command line I'm using:<br>
                          i.cluster group=TM_r15c33@PERMANENT
                          subgroup=AllBands signaturefile=IsoDataSigs2
                          classes=20 sample=5,5 convergence=99.9
                          min_size=10000<br>
                          <br>
                          Is anyone able to get i.cluster to work
                          properly? It would be helpful to know if it's
                          a user problem or a bug.<br>
                          <br>
                          Thanks in advance to anyone who can help.<br>
                          <br>
                          All the best,<br>
                          <br>
                          Ned<br>
_______________________________________________<br>
                          grass-user mailing list<br>
                          <a moz-do-not-send="true"
                            href="mailto:grass-user@lists.osgeo.org">grass-user@lists.osgeo.org</a><br>
                          <a moz-do-not-send="true"
                            href="http://lists.osgeo.org/mailman/listinfo/grass-user">http://lists.osgeo.org/mailman/listinfo/grass-user</a><br>
                        </blockquote>
                      </div>
                      <br>
                    </div>
                  </div>
                  <br>
                  _______________________________________________<br>
                  grass-user mailing list<br>
                  <a moz-do-not-send="true"
                    href="mailto:grass-user@lists.osgeo.org">grass-user@lists.osgeo.org</a><br>
                  <a moz-do-not-send="true"
                    href="http://lists.osgeo.org/mailman/listinfo/grass-user">http://lists.osgeo.org/mailman/listinfo/grass-user</a><br>
                  <br>
                </blockquote>
              </div>
              <br>
            </div>
          </div>
        </div>
      </blockquote>
    </blockquote>
    <br>
  </body>
</html>