<html><head></head><body>You're right, sorry.<br>Any way to improve this, perhaps taking a larger sample to reduce variation?<br>Thanks.<br><br><div class="gmail_quote">Il 3 marzo 2019 05:33:04 CET, Nyall Dawson <nyall.dawson@gmail.com> ha scritto:<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt 0.8ex; border-left: 1px solid rgb(204, 204, 204); padding-left: 1ex;">
<pre class="k9mail">On Sat, 2 Mar 2019 at 22:09, Paolo Cavallini <cavallini@faunalia.it> wrote:<br><br><blockquote class="gmail_quote" style="margin: 0pt 0pt 1ex 0.8ex; border-left: 1px solid #729fcf; padding-left: 1ex;">we also noticed that for large data sets (tens of thousands records)<br>`Pretty breaks` classification behaves strangely: when clicking on<br>Classify repeatedly, classes boundaries change by about 10% all the<br>time, apparently in a random fashion.<br>Unsure whether this is the same or a different issue.<br></blockquote><br>Do you mean pretty breaks or natural breaks (Jenks)? Jenks has been<br>that way forever -- by design (for performance) it takes a random<br>sample of the layer and generates the breaks from that sample only.<br>The random samples is selected anew with each click of classify, which<br>results in the different breaks.<br><br>Nyall<br><br><blockquote class="gmail_quote" style="margin: 0pt 0pt 1ex 0.8ex; border-left: 1px solid #729fcf; padding-left: 1ex;">All the best.<br>--<br>Paolo Cavallini - www.faunalia.eu<br>QGIS.ORG Chair:<br><a href="http://planet.qgis.org/planet/user/28/tag/qgis%20board/">http://planet.qgis.org/planet/user/28/tag/qgis%20board/</a><hr>QGIS-Developer mailing list<br>QGIS-Developer@lists.osgeo.org<br>List info: <a href="https://lists.osgeo.org/mailman/listinfo/qgis-developer">https://lists.osgeo.org/mailman/listinfo/qgis-developer</a><br>Unsubscribe: <a href="https://lists.osgeo.org/mailman/listinfo/qgis-developer">https://lists.osgeo.org/mailman/listinfo/qgis-developer</a><br></blockquote></pre></blockquote></div><br>-- <br>Sorry for being short</body></html>