<div dir="ltr"><div>There are multiple methods for determining the &#39;optimal&#39; number of clusters.<br>The issue is ... which one?<br><br>A statistician was asked for suggestions. In classic &#39;statistician-speak&#39;, he 
rambled on about assorted methods to estimate the optimal number of 
clusters. It was a bit confusing. Esentially, his answer was &#39;try a 
bunch and see what you like.&#39;<br><br></div>NCL is not &#39;R&#39; or Matlab or .... NCL can not provide and support multiple methods. NCL would like to include one optimal cluster function. BUT don&#39;t have the time
 to really look at this issue. However, if you find something you like, 
we will see what we can do.uses 
clustering (&quot;classification&quot;) methods. <br><div><div><br>===<br><br>Several papers, which uses classification/clustering are: <br><br><a href="http://www.cgd.ucar.edu/staff/cdeser/docs/cassou.natlregimes.jclim04.pdf" target="_blank">http://www.cgd.ucar.edu/staff/<wbr>cdeser/docs/cassou.<wbr>natlregimes.jclim04.pdf</a><br></div><div>**See Part 2 (Classification)**<br></div><div><br><a href="http://www.nature.com/nature/journal/v455/n7212/full/nature07286.html" target="_blank">http://www.nature.com/nature/<wbr>journal/v455/n7212/full/<wbr>nature07286.html</a><br><br><a href="ftp://ftp.cerfacs.fr/pub/globc/exchanges/cassou/DCLIM/cassou_et_al_climdyn_regime_ocean_part1.pdf" target="_blank">ftp://ftp.cerfacs.fr/pub/<wbr>globc/exchanges/cassou/DCLIM/<wbr>cassou_et_al_climdyn_regime_<wbr>ocean_part1.pdf</a><br><br><br></div><div>HTH<br></div><div>D<br></div><br><br></div></div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Aug 17, 2016 at 7:02 AM, Carl Schreck <span dir="ltr">&lt;<a href="mailto:cjschrec@ncsu.edu" target="_blank">cjschrec@ncsu.edu</a>&gt;</span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">I&#39;m just starting to play around with the new <a href="http://www.ncl.ucar.edu/Document/Functions/Built-in/kmeans_as136.shtml" target="_blank">kmeans_as136</a> function. I&#39;m very much a novice at clustering and am wondering if anyone has used this function and has guidance on how best to choose the number of clusters. <div><br></div><div>Most of what I&#39;m finding is similar to <a href="https://en.wikipedia.org/wiki/Determining_the_number_of_clusters_in_a_data_set" target="_blank">this wikipedia article</a>. I think the elbow method is the only one that we can do with the output from kmeans_as136, but I&#39;m not exactly sure how to relate ss2 to %variance.</div><div><br></div><div>I&#39;m sure this is simple and I&#39;ve just not read up enough on it, but I figured I&#39;d ask you guys. Thanks!</div><div><br></div><div>Carl<br clear="all"><div><br></div>-- <br><div data-smartmail="gmail_signature"><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><div><div dir="ltr"><br><table style="max-width:100%;border-collapse:collapse;border-spacing:0px;color:rgb(51,51,51);font-size:14px;border:3px solid rgb(170,170,170);font-family:Times;line-height:12px;background-color:transparent" border="0" cellpadding="2" cellspacing="2"><tbody><tr><td height="71" align="center" width="71"><span style="font-size:11px"><span style="font-family:arial,helvetica,sans-serif"><a href="http://www.cicsnc.org/" style="color:rgb(38,58,143);text-decoration:none;font-weight:bold" target="_blank"><img src="http://www.cicsnc.org/assets/images/cicsnc-logo.png" height="93" width="96"></a>  <a href="http://www.researcherid.com/rid/B-8711-2011" target="_blank"><br></a></span></span><br><font size="4"><b style="text-align:start;color:rgb(38,58,143);font-family:arial,helvetica,sans-serif"> <a href="http://www.cyclonecenter.org/" style="color:rgb(38,58,143);text-decoration:none" target="_blank">Cyclone</a></b><a href="http://www.cyclonecenter.org/" style="text-align:start;color:rgb(38,58,143);font-family:arial,helvetica,sans-serif;text-decoration:none" target="_blank">Center.org</a>  </font></td><td valign="top"><span style="font-size:11px"><span style="font-family:arial,helvetica,sans-serif"><b>Carl J. Schreck III, PhD</b><br></span></span><span style="font-size:11px"><span style="font-family:arial,helvetica,sans-serif"><b>Research Scholar / Research </b></span></span><b style="font-family:arial,helvetica,sans-serif;font-size:11px">Assistant Professor</b><span style="font-size:11px"><span style="font-family:arial,helvetica,sans-serif"><br></span></span><a href="http://www.cicsnc.org/" style="font-family:arial,helvetica,sans-serif;font-size:11px;color:rgb(38,58,143);text-decoration:none;font-weight:bold" target="_blank">Cooperative Institute for Climate and Satellites NC (CICS</a><b><a href="http://www.cicsnc.org/" style="font-family:arial,helvetica,sans-serif;font-size:11px;color:rgb(38,58,143);text-decoration:none" target="_blank">-NC)</a> /</b><br><span style="font-size:11px"><span style="font-family:arial,helvetica,sans-serif"><a href="https://meas.sciences.ncsu.edu/" style="color:rgb(38,58,143);text-decoration:none;font-weight:bold" target="_blank">Department of Marine, Earth, and Atmospheric Sciences (MEAS)</a> <br><a href="http://ncsu.edu/" style="color:rgb(38,58,143);text-decoration:none;font-weight:bold" target="_blank">North Carolina State University</a><br><a href="http://ncei.noaa.gov/" style="color:rgb(38,58,143);text-decoration:none;font-weight:bold" target="_blank">NOAA National Centers for Environmental Information (NCEI)</a><br>151 Patton Ave, Asheville, NC 28801<br>e: </span></span><a href="mailto:carl_schreck@ncsu.edu" style="color:rgb(38,58,143);font-family:arial,helvetica,sans-serif;font-size:11px;text-decoration:none;font-weight:bold" target="_blank">carl_schreck@ncsu.edu</a><br style="font-family:arial,helvetica,sans-serif;font-size:11px"><span style="font-size:11px"><span style="font-family:arial,helvetica,sans-serif">o: <a href="tel:%2B1%20828%20257%203140" value="+18282573140" target="_blank">+1 828 257 3140</a><br></span></span><span style="font-family:arial,helvetica,sans-serif;font-size:11px">c: <a href="tel:%2B1%20828%20484%201702" value="+18284841702" target="_blank">+1 828 484 1702</a><br></span><a href="http://scholar.google.com/citations?hl=en&amp;user=th8ONEcAAAAJ&amp;view_op=list_works&amp;sortby=pubdate" style="color:rgb(38,58,143);font-family:arial,helvetica,sans-serif;font-size:11px;text-decoration:none;font-weight:bold" target="_blank">Publications</a><br><a href="http://monitor.cicsnc.org/mjo/" style="color:rgb(38,58,143);font-family:arial,helvetica,sans-serif;font-size:11px;text-decoration:none;font-weight:bold" target="_blank">monitor.cicsnc.org/mjo</a><br style="font-family:arial,helvetica,sans-serif;font-size:11px"></td></tr></tbody></table></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div></div>
</div></div>
<br>______________________________<wbr>_________________<br>
ncl-talk mailing list<br>
<a href="mailto:ncl-talk@ucar.edu">ncl-talk@ucar.edu</a><br>
List instructions, subscriber options, unsubscribe:<br>
<a href="http://mailman.ucar.edu/mailman/listinfo/ncl-talk" rel="noreferrer" target="_blank">http://mailman.ucar.edu/<wbr>mailman/listinfo/ncl-talk</a><br>
<br></blockquote></div><br></div>