<?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 Partition On Proportion(percentage) Responses in Discussions</title>
    <link>https://community.jmp.com/t5/Discussions/Partition-On-Proportion-percentage-Responses/m-p/888644#M105093</link>
    <description>&lt;P&gt;Dear JMP Community,&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I have tried to find similar post but so far none is addressing the question I have.&lt;/P&gt;
&lt;P&gt;I attached here also dummy data.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I am doing data mining to profile and quantify the factors vs responses effect.&lt;/P&gt;
&lt;P&gt;To put it simple, I have 2 factors (categorical with 2 levels) and 1 response (yield loss) which is a percentage.&lt;/P&gt;
&lt;P&gt;We record each day's yield loss vs what 2 factors level we are using on that day.&lt;/P&gt;
&lt;P&gt;For each day, our quantity produce is different.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The modeling technique I'm using is Partition.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Partition directly calculate Mean percentage as show in the output below.&lt;/P&gt;
&lt;P&gt;As each day the quantity we produce is different, it is not accurate that Partition will simply calculate the Mean of the Yield Loss%.&lt;/P&gt;
&lt;P&gt;&lt;span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="Picture1.png" style="width: 489px;"&gt;&lt;img src="https://community.jmp.com/t5/image/serverpage/image-id/78710iDA6CC586938026D0/image-dimensions/489x353?v=v2" width="489" height="353" role="button" title="Picture1.png" alt="Picture1.png" /&gt;&lt;/span&gt;&lt;/P&gt;
&lt;P&gt;My question is:&lt;/P&gt;
&lt;P&gt;1. How do we calculate the Mean weighted percentage in this case using Partition? I'm seeing an option in the Partition Platform for "Weight", but I'm not sure how to utilize that.&lt;/P&gt;
&lt;P&gt;2. Is there a feature in Partition we could use Median instead of Mean? (since the yield loss distribution are generally skewed).&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Thanks for your advice,&lt;/P&gt;
&lt;P&gt;Zappy&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
    <pubDate>Wed, 23 Jul 2025 03:59:23 GMT</pubDate>
    <dc:creator>Zappy</dc:creator>
    <dc:date>2025-07-23T03:59:23Z</dc:date>
    <item>
      <title>Partition On Proportion(percentage) Responses</title>
      <link>https://community.jmp.com/t5/Discussions/Partition-On-Proportion-percentage-Responses/m-p/888644#M105093</link>
      <description>&lt;P&gt;Dear JMP Community,&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I have tried to find similar post but so far none is addressing the question I have.&lt;/P&gt;
&lt;P&gt;I attached here also dummy data.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;I am doing data mining to profile and quantify the factors vs responses effect.&lt;/P&gt;
&lt;P&gt;To put it simple, I have 2 factors (categorical with 2 levels) and 1 response (yield loss) which is a percentage.&lt;/P&gt;
&lt;P&gt;We record each day's yield loss vs what 2 factors level we are using on that day.&lt;/P&gt;
&lt;P&gt;For each day, our quantity produce is different.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;The modeling technique I'm using is Partition.&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Partition directly calculate Mean percentage as show in the output below.&lt;/P&gt;
&lt;P&gt;As each day the quantity we produce is different, it is not accurate that Partition will simply calculate the Mean of the Yield Loss%.&lt;/P&gt;
&lt;P&gt;&lt;span class="lia-inline-image-display-wrapper lia-image-align-inline" image-alt="Picture1.png" style="width: 489px;"&gt;&lt;img src="https://community.jmp.com/t5/image/serverpage/image-id/78710iDA6CC586938026D0/image-dimensions/489x353?v=v2" width="489" height="353" role="button" title="Picture1.png" alt="Picture1.png" /&gt;&lt;/span&gt;&lt;/P&gt;
&lt;P&gt;My question is:&lt;/P&gt;
&lt;P&gt;1. How do we calculate the Mean weighted percentage in this case using Partition? I'm seeing an option in the Partition Platform for "Weight", but I'm not sure how to utilize that.&lt;/P&gt;
&lt;P&gt;2. Is there a feature in Partition we could use Median instead of Mean? (since the yield loss distribution are generally skewed).&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;
&lt;P&gt;Thanks for your advice,&lt;/P&gt;
&lt;P&gt;Zappy&lt;/P&gt;
&lt;P&gt;&amp;nbsp;&lt;/P&gt;</description>
      <pubDate>Wed, 23 Jul 2025 03:59:23 GMT</pubDate>
      <guid>https://community.jmp.com/t5/Discussions/Partition-On-Proportion-percentage-Responses/m-p/888644#M105093</guid>
      <dc:creator>Zappy</dc:creator>
      <dc:date>2025-07-23T03:59:23Z</dc:date>
    </item>
    <item>
      <title>Re: Partition On Proportion(percentage) Responses</title>
      <link>https://community.jmp.com/t5/Discussions/Partition-On-Proportion-percentage-Responses/m-p/888699#M105103</link>
      <description>&lt;P&gt;Hi &lt;a href="https://community.jmp.com/t5/user/viewprofilepage/user-id/8769"&gt;@Zappy&lt;/a&gt;,&lt;BR /&gt;&lt;BR /&gt;You could maybe add a column with the number of batches/quantity per day, and use this information as Frequency information when launching the Partition platform : &lt;A href="https://www.jmp.com/support/help/en/18.2/#page/jmp/launch-the-partition-platform.shtml#ww1276002" target="_blank" rel="noopener"&gt;https://www.jmp.com/support/help/en/18.2/#page/jmp/launch-the-partition-platform.shtml#ww1276002&lt;/A&gt;&lt;BR /&gt;&lt;BR /&gt;The Decision Tree do not create splits directly based on mean values, see &lt;A href="https://www.jmp.com/support/help/en/18.2/#page/jmp/statistical-details-for-the-splitting-criterion.shtml#ww1277127" target="_blank" rel="noopener"&gt;https://www.jmp.com/support/help/en/18.2/#page/jmp/statistical-details-for-the-splitting-criterion.shtml#ww1277127&lt;/A&gt; for more info. So I think you could still get valuable insights using the platform directly.&lt;BR /&gt;If not, you could maybe try transforming your data, maybe by binning your raw data into classes of values first or log-transform your raw data (if the distribution is approximately log-normal), before partitioning with Decision Tree ?&lt;BR /&gt;&lt;BR /&gt;Hope this answer will help you,&lt;/P&gt;</description>
      <pubDate>Thu, 24 Jul 2025 08:00:35 GMT</pubDate>
      <guid>https://community.jmp.com/t5/Discussions/Partition-On-Proportion-percentage-Responses/m-p/888699#M105103</guid>
      <dc:creator>Victor_G</dc:creator>
      <dc:date>2025-07-24T08:00:35Z</dc:date>
    </item>
  </channel>
</rss>

