<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: Microsoft Chart Advisor</title>
	<atom:link href="http://peltiertech.com/WordPress/microsoft-chart-advisor/feed/" rel="self" type="application/rss+xml" />
	<link>http://peltiertech.com/WordPress/microsoft-chart-advisor/</link>
	<description>Peltier Tech Excel Charts and Programming Blog</description>
	<lastBuildDate>Fri, 10 Feb 2012 07:51:10 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.1.2</generator>
<xhtml:meta xmlns:xhtml="http://www.w3.org/1999/xhtml" name="robots" content="noindex" />
	<item>
		<title>By: Dynamic Diagrams : Information Design Watch : Microsoft Chart Advisor&#8212;Consider the Source</title>
		<link>http://peltiertech.com/WordPress/microsoft-chart-advisor/comment-page-1/#comment-5936</link>
		<dc:creator>Dynamic Diagrams : Information Design Watch : Microsoft Chart Advisor&#8212;Consider the Source</dc:creator>
		<pubDate>Mon, 03 Nov 2008 20:51:23 +0000</pubDate>
		<guid isPermaLink="false">http://peltiertech.com/WordPress/?p=375#comment-5936</guid>
		<description>[...] Tim Mays reported that Chart Advisor ignored a whole column of source data and then (not surprisingly) recommended the wrong chart type. Excel guru Jon Peltier didn&#8217;t even get that far. [...]</description>
		<content:encoded><![CDATA[<p>[...] Tim Mays reported that Chart Advisor ignored a whole column of source data and then (not surprisingly) recommended the wrong chart type. Excel guru Jon Peltier didn&#8217;t even get that far. [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jon Peltier</title>
		<link>http://peltiertech.com/WordPress/microsoft-chart-advisor/comment-page-1/#comment-3280</link>
		<dc:creator>Jon Peltier</dc:creator>
		<pubDate>Fri, 29 Aug 2008 17:48:15 +0000</pubDate>
		<guid isPermaLink="false">http://peltiertech.com/WordPress/?p=375#comment-3280</guid>
		<description>Hi Dave -

Thanks for the insights into your rules database. The number of lines doesn&#039;t scare me, just figuring out syntax and so forth might be a challenge. Looking it over will allow me some understanding of your thought processes and how you chose to implement certain rules.

I understand that development works in funny ways, and that in the Chart Advisor, Excel is the recipient of code originally not intended for Excel. The initial hiccups lead to an underwhelming first impression. My second impression was that the utility was pretty solid, and that I&#039;d have more issues with your rules and their implementations than in the way the utility actually works.</description>
		<content:encoded><![CDATA[<p>Hi Dave -</p>
<p>Thanks for the insights into your rules database. The number of lines doesn&#8217;t scare me, just figuring out syntax and so forth might be a challenge. Looking it over will allow me some understanding of your thought processes and how you chose to implement certain rules.</p>
<p>I understand that development works in funny ways, and that in the Chart Advisor, Excel is the recipient of code originally not intended for Excel. The initial hiccups lead to an underwhelming first impression. My second impression was that the utility was pretty solid, and that I&#8217;d have more issues with your rules and their implementations than in the way the utility actually works.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Dave Lee</title>
		<link>http://peltiertech.com/WordPress/microsoft-chart-advisor/comment-page-1/#comment-3275</link>
		<dc:creator>Dave Lee</dc:creator>
		<pubDate>Fri, 29 Aug 2008 15:40:42 +0000</pubDate>
		<guid isPermaLink="false">http://peltiertech.com/WordPress/?p=375#comment-3275</guid>
		<description>VBA vs .NET
This project began as a generic analysis core, with data sources and a custom presentation layer.  We developed it in .NET C#, with Excel as the first target.  VBA is great, but obviously with this approach, VBA wasn&#039;t an option for us.  

VSTO 2008 has made deployment much easier than the VSTO 2005 release.  However, we stuck with the 2005 release to allow the broadest number of people to use our prototype. We had some hiccups in our deployment, but it seems to be ironed out now.

Thanks

Dave Lee
Chart Advisor from Microsoft Office Labs
Developer</description>
		<content:encoded><![CDATA[<p>VBA vs .NET<br />
This project began as a generic analysis core, with data sources and a custom presentation layer.  We developed it in .NET C#, with Excel as the first target.  VBA is great, but obviously with this approach, VBA wasn&#8217;t an option for us.  </p>
<p>VSTO 2008 has made deployment much easier than the VSTO 2005 release.  However, we stuck with the 2005 release to allow the broadest number of people to use our prototype. We had some hiccups in our deployment, but it seems to be ironed out now.</p>
<p>Thanks</p>
<p>Dave Lee<br />
Chart Advisor from Microsoft Office Labs<br />
Developer</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Dave Lee</title>
		<link>http://peltiertech.com/WordPress/microsoft-chart-advisor/comment-page-1/#comment-3273</link>
		<dc:creator>Dave Lee</dc:creator>
		<pubDate>Fri, 29 Aug 2008 15:31:51 +0000</pubDate>
		<guid isPermaLink="false">http://peltiertech.com/WordPress/?p=375#comment-3273</guid>
		<description>Hi Jon.  Thanks for your comments.  

Our rules files is around 5000 lines.  It is called ChartAdvisor.Rules.xml, and it is installed to your AppData directory (you&#039;ll need to search since the location depends upon your OS and configuration).  In the tooltips, we show you only the scoring rules which evaluated to true.  We also have rules for data mapping (figuring out which Excel series should be mapped to chart axis), and formatting rules, which allow things like turning on the gummy styles only when there are a few bars on a column chart, or moving legends in to the slices on a pie chart when the smallest slice is &gt; 5%. 

btw, I wouldn&#039;t spend time trying to change the rules file -- the syntax is fairly rigid and modifications by hand can cause it to be easily broken, so we overwrite the file when we start up.

The folks on the immediate development team here are not charting experts, but we&#039;ve been talking with SME&#039;s around here for guidance in improving the rules, and we&#039;d love to open a dialog with you as well.

Thanks 

Dave Lee
Chart Advisor from Microsoft Office Labs
Developer</description>
		<content:encoded><![CDATA[<p>Hi Jon.  Thanks for your comments.  </p>
<p>Our rules files is around 5000 lines.  It is called ChartAdvisor.Rules.xml, and it is installed to your AppData directory (you&#8217;ll need to search since the location depends upon your OS and configuration).  In the tooltips, we show you only the scoring rules which evaluated to true.  We also have rules for data mapping (figuring out which Excel series should be mapped to chart axis), and formatting rules, which allow things like turning on the gummy styles only when there are a few bars on a column chart, or moving legends in to the slices on a pie chart when the smallest slice is &gt; 5%. </p>
<p>btw, I wouldn&#8217;t spend time trying to change the rules file &#8212; the syntax is fairly rigid and modifications by hand can cause it to be easily broken, so we overwrite the file when we start up.</p>
<p>The folks on the immediate development team here are not charting experts, but we&#8217;ve been talking with SME&#8217;s around here for guidance in improving the rules, and we&#8217;d love to open a dialog with you as well.</p>
<p>Thanks </p>
<p>Dave Lee<br />
Chart Advisor from Microsoft Office Labs<br />
Developer</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Tim Mayes</title>
		<link>http://peltiertech.com/WordPress/microsoft-chart-advisor/comment-page-1/#comment-3264</link>
		<dc:creator>Tim Mayes</dc:creator>
		<pubDate>Thu, 28 Aug 2008 22:04:16 +0000</pubDate>
		<guid isPermaLink="false">http://peltiertech.com/WordPress/?p=375#comment-3264</guid>
		<description>Jon,

This update seems to have removed all of the eye-candy from the charts. Also, the chart scores are now much higher (for my original data set) than they were previously, and I can no longer see how the scores are determined. On the plus side, it does give more choices than I previously got.

Oh, I also now get a message saying that it is only a prototype and hasn&#039;t been tuned for performance. Interestingly, the updated version seems slower than the original.</description>
		<content:encoded><![CDATA[<p>Jon,</p>
<p>This update seems to have removed all of the eye-candy from the charts. Also, the chart scores are now much higher (for my original data set) than they were previously, and I can no longer see how the scores are determined. On the plus side, it does give more choices than I previously got.</p>
<p>Oh, I also now get a message saying that it is only a prototype and hasn&#8217;t been tuned for performance. Interestingly, the updated version seems slower than the original.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Jon Peltier</title>
		<link>http://peltiertech.com/WordPress/microsoft-chart-advisor/comment-page-1/#comment-3258</link>
		<dc:creator>Jon Peltier</dc:creator>
		<pubDate>Thu, 28 Aug 2008 19:40:23 +0000</pubDate>
		<guid isPermaLink="false">http://peltiertech.com/WordPress/?p=375#comment-3258</guid>
		<description>&lt;strong&gt;Update&lt;/strong&gt;

I&#039;ll follow up with a &quot;formal&quot; blog entry, but I thought I&#039;d check in. MS Labs have fixed whatever the glitches were with the installation of the first version of the chart advisor. I was able to install the add-in fine this time. I still think any addin that requires PIAs is over the top, but that&#039;s my opinion, and you&#039;re entitled to it because you&#039;re reading my blog.

Anyway, it installed nicely, and it sits there next to the chart group on the Insert tab. Select data and click the button, it thinks a moment, and comes up with a set of selections. It is pretty slick, and I managed to crash it with the first data set I selected (I wasn&#039;t trying, I promise!). I selected a 3x3 range of blank cells, and I was told it was an unexpected situation. Okay, that&#039;s fair, the instructions say to select data then click, and 3/4 of your users will use it that way. Of course, the utility had turned off screen updating, which it probably doesn&#039;t need to do, then didn&#039;t restore it when I canceled.

So manually I restored screen updating and tried a few ranges. The interface is pretty slick, and I like the fact that it provides the reasoning behind its chart selections. So many points because of X, so many because of Y, etc. 

I&#039;ll hold off on a more detailed review until I have time to do it justice, but it made a reasonably good first impression. I made no judgment yet about the suggested chart types, but I will. I noticed some things that deserve mention:

&lt;ul style=&quot;margin-left : 20px&quot;&gt;&lt;li&gt;The Advisor doesn&#039;t seem to know the trick of leaving the top left cell(s) blank to indicate series names and categories range.&lt;/li&gt;

&lt;li&gt;The Advisor only seems to want to make charts with series in columns.&lt;/li&gt;

&lt;li&gt;The advisor knows when the data range contains no numerical values (but misses all blanks).&lt;/li&gt;

&lt;li&gt;The Advisor runs many of the checks of category values as my custom apps to: are they text, are they dates, are they years?&lt;/li&gt;

&lt;li&gt;The Advisor seems to evaluate some of the properties of the data, including number of distinct values, spread of the values, etc.&lt;/li&gt;&lt;/ul&gt;

Given that it contains a set of rules somewhere, I wonder if eventually it would be possible to rearrange the rules, apply custom values to the rules, even create your own rules. This would make it a very cool application.</description>
		<content:encoded><![CDATA[<p><strong>Update</strong></p>
<p>I&#8217;ll follow up with a &#8220;formal&#8221; blog entry, but I thought I&#8217;d check in. MS Labs have fixed whatever the glitches were with the installation of the first version of the chart advisor. I was able to install the add-in fine this time. I still think any addin that requires PIAs is over the top, but that&#8217;s my opinion, and you&#8217;re entitled to it because you&#8217;re reading my blog.</p>
<p>Anyway, it installed nicely, and it sits there next to the chart group on the Insert tab. Select data and click the button, it thinks a moment, and comes up with a set of selections. It is pretty slick, and I managed to crash it with the first data set I selected (I wasn&#8217;t trying, I promise!). I selected a 3&#215;3 range of blank cells, and I was told it was an unexpected situation. Okay, that&#8217;s fair, the instructions say to select data then click, and 3/4 of your users will use it that way. Of course, the utility had turned off screen updating, which it probably doesn&#8217;t need to do, then didn&#8217;t restore it when I canceled.</p>
<p>So manually I restored screen updating and tried a few ranges. The interface is pretty slick, and I like the fact that it provides the reasoning behind its chart selections. So many points because of X, so many because of Y, etc. </p>
<p>I&#8217;ll hold off on a more detailed review until I have time to do it justice, but it made a reasonably good first impression. I made no judgment yet about the suggested chart types, but I will. I noticed some things that deserve mention:</p>
<ul style="margin-left : 20px">
<li>The Advisor doesn&#8217;t seem to know the trick of leaving the top left cell(s) blank to indicate series names and categories range.</li>
<li>The Advisor only seems to want to make charts with series in columns.</li>
<li>The advisor knows when the data range contains no numerical values (but misses all blanks).</li>
<li>The Advisor runs many of the checks of category values as my custom apps to: are they text, are they dates, are they years?</li>
<li>The Advisor seems to evaluate some of the properties of the data, including number of distinct values, spread of the values, etc.</li>
</ul>
<p>Given that it contains a set of rules somewhere, I wonder if eventually it would be possible to rearrange the rules, apply custom values to the rules, even create your own rules. This would make it a very cool application.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

