<div dir="ltr"><div><div><div><div><div><div>The documentation for eof2data needs to be cleared up.<br><br> evecv = eofunc_Wrap (x,neval,False)<br> evecv_ts = eofunc_ts_Wrap (x,evecv,False)<br><br> x0 = eof2data (evecv,evecv_ts)<br><br></div>The returned x0 will have the correct variance but will be offset from the expected result.<br><br></div>To return the correct values the 'evecv_ts' must have each evecv_ts@ts_mean(n) added back<br></div>*prior* to use by eof2data.<br><br> <br> do n=0,neval-1<br> evecv_ts(n,:) = evecv_ts(n,:) + evecv_ts@ts_mean(n) ; add component means back in<br> end do<br> x1 = eof2data (evecv,evecv_ts)<br><br></div>Attached is a script which uses data drom a book. NCL matches the book results exactly.<br><br></div>Regards<br></div>D<br><div><div><div><div class="gmail_extra"><br><div class="gmail_quote">On Wed, Nov 12, 2014 at 8:31 AM, Xia, Geng <span dir="ltr"><<a href="mailto:gxia@albany.edu" target="_blank">gxia@albany.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div dir="ltr">
<div style="font-size:12pt;color:#000000;background-color:#ffffff;font-family:Calibri,Arial,Helvetica,sans-serif">
<p>Hi all<br>
</p>
<p><br>
</p>
<p>I send out message yesterday, reporting on a problem with EOF function in ncl. Thanks to many others who helps me, It<br>
</p>
<p>turns out that the EOF function miscalculate the first eof mode (the mean). I did an experiment by removing the mean<span style="font-size:12pt"> from the dataset, then I perform the EOF function and reco</span><span style="font-size:12pt">nstruct the data.
The result is perfect after that. </span></p>
<p><span style="font-size:12pt"><br>
</span></p>
<p><span style="font-family:Calibri,Arial,Helvetica,sans-serif;font-size:16px;background-color:rgb(255,255,255)">This problem won't matter if people is working with anomalies data. However, It will be an issue if people use the EOF as a filter. For example,
if you use EOF to filter the high frequency noise from data set and reconstruct the filtered data. It will cause uncertainty. </span><br>
</p>
<p><br>
</p>
<p>So hopefully, someone who is managing the EOF coding can see this message. <span style="font-size:12pt">The function I think is responsi</span><span style="font-size:12pt">b</span><span style="font-size:12pt">le for the problem is eofunc_ts and eofunc</span><span style="font-size:12pt">.</span></p>
<p><br>
</p>
<p>Best <span class="HOEnZb"><font color="#888888"><br>
</font></span></p><span class="HOEnZb"><font color="#888888">
<p>Geng Xia<br>
</p>
<p><br>
</p>
</font></span></div>
</div>
<br>_______________________________________________<br>
ncl-talk mailing list<br>
List instructions, subscriber options, unsubscribe:<br>
<a href="http://mailman.ucar.edu/mailman/listinfo/ncl-talk" target="_blank">http://mailman.ucar.edu/mailman/listinfo/ncl-talk</a><br>
<br></blockquote></div><br></div></div></div></div></div>