<?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: Really odd behavior in QlikView</title>
    <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385109#M143844</link>
    <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;If it's order by the same way - you get the same result &lt;IMG src="https://community.qlik.com/legacyfs/online/emoticons/happy.png" /&gt;&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;I guess, I add some fields so that I can avoid interrecord functions.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Whatever, your expression does not imply user selections.&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
    <pubDate>Thu, 06 Sep 2012 17:48:39 GMT</pubDate>
    <dc:creator>whiteline</dc:creator>
    <dc:date>2012-09-06T17:48:39Z</dc:date>
    <item>
      <title>Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385104#M143839</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;Hi all.&lt;/P&gt;&lt;P&gt;I have an expression which calculates days depending if the previous 3 weeks are there in the data. For example, if 2012W23 to 2012W27 is there for every seller, the result is 28 days for every week, if there is one week missing, i.e. 2012W25 the result is 21 days, if there are 2 weeks missing i.e. 2012W23 and 2012W25, the result is 14 days, if all 3 previous weeks are missing, the result is 7 days.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;This seems to be working well in a straight table. I take the average of rows as a total, and the result is quite fine. it does not matter which weeks are missing, the result is as expected.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;But I need to present this average of rows as one number in a text box, so the obvious choice was to aggr the expression over seller and week, in order to mimic the chart. See attached file.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN style="color: #339966;"&gt;If all weeks are present, the aggr gives the same result as the chart.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN style="color: #339966;"&gt;If week 2012W12 for seller a is omitted, the aggr gives the same result as the chart.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN style="color: #339966;"&gt;If week 2012W09 for seller a is omitted and seller b has data up to week 2012W10, the aggr gives the same result as the chart.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;SPAN style="color: #ff0000;"&gt;If week 2012W09 for seller a is omitted and seller b has weeks present over week 2012W10 for example 2012W11, the result is different.&lt;/SPAN&gt;&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;You can try these scnarios in the attached file.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;I only want to aggr the values the chart is showing, I cannot understand why it is related to data. The chart is working, the aggr is not, depending on the data.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Any ideas?&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Thu, 06 Sep 2012 15:19:52 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385104#M143839</guid>
      <dc:creator>giakoum</dc:creator>
      <dc:date>2012-09-06T15:19:52Z</dc:date>
    </item>
    <item>
      <title>Re: Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385105#M143840</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;I created a list box with the text box expression to see where the difference comes from. It seems that when a week, for example 2012W09 is missing from Seller a, seller b has wrong results from 2012W09 and onwards. I cannot understand why this happens in the aggr only and not the chart itself :&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;&lt;IMG __jive_id="21451" __jive_id="21451" class="jive-image-thumbnail jive-image" alt="2012-09-06_1848.png" src="https://community.qlik.com/legacyfs/online/21451_2012-09-06_1848.png" width="450" /&gt;&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Thu, 06 Sep 2012 15:53:18 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385105#M143840</guid>
      <dc:creator>giakoum</dc:creator>
      <dc:date>2012-09-06T15:53:18Z</dc:date>
    </item>
    <item>
      <title>Re: Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385106#M143841</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;I guess, the problem is with your expression not with aggr.&lt;/P&gt;&lt;P&gt;You use inter record functions with aggr.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Why do you expect that tha data representation and &lt;SPAN style="text-decoration: underline;"&gt;sorting&lt;/SPAN&gt; for (inside) aggr is the same as in your chart ?&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Thu, 06 Sep 2012 16:52:05 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385106#M143841</guid>
      <dc:creator>whiteline</dc:creator>
      <dc:date>2012-09-06T16:52:05Z</dc:date>
    </item>
    <item>
      <title>Re: Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385107#M143842</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;Try 'load order' as sort order and you'll see the difference.&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Thu, 06 Sep 2012 16:54:42 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385107#M143842</guid>
      <dc:creator>whiteline</dc:creator>
      <dc:date>2012-09-06T16:54:42Z</dc:date>
    </item>
    <item>
      <title>Re: Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385108#M143843</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;True but that is what aggr function is doing. It is ordering the same way, as you can see from the comparison between the chart and the list box.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;How would you solve this?&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Thu, 06 Sep 2012 17:12:58 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385108#M143843</guid>
      <dc:creator>giakoum</dc:creator>
      <dc:date>2012-09-06T17:12:58Z</dc:date>
    </item>
    <item>
      <title>Re: Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385109#M143844</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;If it's order by the same way - you get the same result &lt;IMG src="https://community.qlik.com/legacyfs/online/emoticons/happy.png" /&gt;&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;I guess, I add some fields so that I can avoid interrecord functions.&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Whatever, your expression does not imply user selections.&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Thu, 06 Sep 2012 17:48:39 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385109#M143844</guid>
      <dc:creator>whiteline</dc:creator>
      <dc:date>2012-09-06T17:48:39Z</dc:date>
    </item>
    <item>
      <title>Re: Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385110#M143845</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;Well, I was first pissed with whitelines's answer, but then I thought I look into it a little. And yes, whiteline was right, aggr takes load order by default, it always does, and since 2012W09 was missing from the first seller the order was as follows :&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;....&lt;/P&gt;&lt;P&gt;2012W08&lt;/P&gt;&lt;P&gt;2012W09&lt;/P&gt;&lt;P&gt;2012W11&lt;/P&gt;&lt;P&gt;2012W12&lt;/P&gt;&lt;P&gt;2012W10 (since this week was only found for seller b!!)&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;So I simply loaded the weeks seperately in the desired order and now everything works fine! See attached file. A simple list box with weeks ordered by load revealed the schocking, but to be expected truth!&lt;/P&gt;&lt;P&gt;&lt;/P&gt;&lt;P&gt;Sorry whiteline for all the nasty thoughts I made &lt;IMG src="https://community.qlik.com/legacyfs/online/emoticons/happy.png" /&gt;, you were really helpfull!&lt;/P&gt;&lt;P&gt;Thank you.&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Thu, 06 Sep 2012 18:35:13 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385110#M143845</guid>
      <dc:creator>giakoum</dc:creator>
      <dc:date>2012-09-06T18:35:13Z</dc:date>
    </item>
    <item>
      <title>Re: Really odd behavior</title>
      <link>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385111#M143846</link>
      <description>&lt;HTML&gt;&lt;HEAD&gt;&lt;/HEAD&gt;&lt;BODY&gt;&lt;P&gt;It's ok. Never mind. I think its normal to be in doubt about comunity comments.&lt;/P&gt;&lt;P&gt;At least you've tried it and I've found something useful too.&lt;/P&gt;&lt;/BODY&gt;&lt;/HTML&gt;</description>
      <pubDate>Fri, 07 Sep 2012 05:42:17 GMT</pubDate>
      <guid>https://community.qlik.com/t5/QlikView/Really-odd-behavior/m-p/385111#M143846</guid>
      <dc:creator>whiteline</dc:creator>
      <dc:date>2012-09-07T05:42:17Z</dc:date>
    </item>
  </channel>
</rss>

