https://sourceforge.net/p/python-cluster/patches/2006-08-05T19:29:03ZRecent changes to patches2006-08-05T19:29:03Z2006-08-05T19:29:03Zajaksuhttps://sourceforge.net/u/ajaksu2/https://sourceforge.netb3cdd6d2b90814e0c4acaf7730f30f78c2f923c2

<div class="markdown_content"><p>Hi Michel,</p> <p>Pretty useful project you have here, I haven't read the<br /> clustering algorithms yet but have used it (with<br /> ecological similarity indexes among samples!).<br /> Hierarchical clustering makes that a LOT more useful,<br /> so thank you very much :)</p> <p>This patch adds a topological output to<br /> Cluster(.topology()) and a respective method to<br /> BaseClusterMethod(.topo()). It also changes .data()<br /> behavior to avoid the .data()[0] idiom.</p> <p>Feeding the output of .topo() to<br /> <a href="http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/139422" rel="nofollow">http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/139422</a><br /> gives (e.g.):</p> <p>9 -----+ <br /> |--+<br /> 8 -----+ |<br /> |--<br /> # --+ |<br /> |--+ |<br /> w --+ | |<br /> |--+<br /> e -----+</p> <p>Regarding the .data() change:<br /> * I think it'd be better to keep two different<br /> datasets (.raw and .data?), because both states (flat<br /> and clustered) are pretty different beasts.<br /> * Perhaps having .data() "contents" stored as<br /> .data would make things easier (in a "why call a method<br /> with no parameters, that returns an object" way :)).<br /> Making .data a property would keep current behavior and<br /> require less ()s :)</p> <p>I'm working on an advanced version of the topological<br /> view and on a derivative of that printing recipe for my<br /> own needs, but will forward them here when done.</p> <p>Thanks again,<br /> Daniel</p></div>