http://blog.ouseful.info/2012/05/02/doodling-with-a-conversation-or-retweet-data-sketch-around-lak12/ Trying to find useful things to do with emerging technologies in open education Thu, 23 May 2013 14:40:45 +0000 hourly 1 http://wordpress.com/ http://blog.ouseful.info/2012/05/02/doodling-with-a-conversation-or-retweet-data-sketch-around-lak12/#comment-22010 Thu, 03 May 2012 00:07:50 +0000 http://blog.ouseful.info/?p=7718#comment-22010 Ah, I see the problem now. In that case maybe, if you’ve got a (dis)similarity matrix, just multiply the rows and columns relating to acqua and orange by some arbitrarily huge positive and negative constant? Again, don’t know if this makes sense in the context of your df/plotting interests

]]> http://blog.ouseful.info/2012/05/02/doodling-with-a-conversation-or-retweet-data-sketch-around-lak12/#comment-22009 Wed, 02 May 2012 23:47:50 +0000 http://blog.ouseful.info/?p=7718#comment-22009 Hi Chris: thanks for the suggestion. I think I did one of those in the LAK12 summary doc I linked to? What I specifically was trying to do was retain some sense of the time axis…

]]> http://blog.ouseful.info/2012/05/02/doodling-with-a-conversation-or-retweet-data-sketch-around-lak12/#comment-22008 Wed, 02 May 2012 23:43:02 +0000 http://blog.ouseful.info/?p=7718#comment-22008 Maybe (a) construct a pairwise similarity matrix where matrix entries (x,y) are retweets of x by y, then use the R package seriation to help with ordering? Bit difficult to tell how to do it with your data structure.

]]>