Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-11-04

| Channels | #opentreeoflife index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
02:36 towodo joined #opentreeoflife
13:01 towodo joined #opentreeoflife
13:53 jimallman joined #opentreeoflife
13:59 jimallman_ joined #opentreeoflife
15:03 kcranstn joined #opentreeoflife
15:20 kcranstn joined #opentreeoflife
15:23 kcranstn joined #opentreeoflife
15:32 mtholder joined #opentreeoflife
15:33 kcranstn joined #opentreeoflife
15:33 mtholder jimallman do you mind if I deploy a branch of opentree to devtree? (mock up of the stacked barcharts discussed on the call yesterday)
15:34 jimallman please do!
15:34 kcranstn yes!
15:34 jimallman mtholder:^
15:34 mtholder ok.
15:35 jimallman fyi - i’ve removed the sidebar nav for About pages (in a branch cleverly called remove-sidebar-nav), if you want to merge that first
15:35 mtholder OK will do.
15:39 mtholder OK. No doubt you'll want to change the colors and do things like make sure that the legend does not block out the graph, but otu-stats-graph branch is deployed at: http://devtree.opentreeoflife.org/about/otu_statistics
15:41 mtholder it needs some more explanation, but for each synth date there are 2 stacked bars: one for where the OTT OTUs fall out in terms of: synthetic tree, nominated studies, and other studies, not in any study
15:42 mtholder and the second bar is the # of OTUs in the studies in the synthetic tree, the # in synthesis but no studies, and otus that are pruned during synthesis (but in OTT)
15:42 mtholder I did not have all of the data, so some numbers are made up (noted by the red text on the page).
16:12 scrollback joined #opentreeoflife
16:15 towodo jimallman, I understand why you want the stats script to be responsible for composing the human-visible label, but I’m not sure it’s the right division of labor
16:15 towodo it’s a model/view situation
16:16 jimallman towodo: understood. i didn’t ask for this, it just turned out that Peter was giving us nice labels, so I thought i’d try showing them.
16:16 jimallman happy to code for these explicitly. (though it’s interesting that the labels have already changed in just two dated profiles)
16:17 towodo the UI is the expert at UI, and it will need to know semantics anyhow in order to decide ordering, graphical presentation, etc.
16:17 towodo at some point we should freeze the label names
16:18 jimallman agreed. normally we’d have terse property names in the JSON: profileDate, totalOTUs, etc. but I can work with the current properties, which might be nice for someone working directly with the JSON files. (which is a minimal UI too, i guess)
16:18 towodo wait, doesn’t the json have both short names, and long labels? … let me look, I guess it’s in my email stream …
16:19 kcranstn joined #opentreeoflife
16:19 towodo oh. i guess it doesn’t.
16:20 towodo let’s fix that
16:20 towodo I’ll email this irc pointer to peter
17:08 mtholder joined #opentreeoflife
17:50 pmidford2 joined #opentreeoflife
18:19 kcranstn Thinking about conflict, and wondering if we could come up with a simple measure of conflict, bin that and then label nodes in the tree as low, medium, high conflict areas
18:22 mtholder joined #opentreeoflife
18:28 towodo any node partitions the set of its tree’s OTUs into an ingroup and an outgroup. so for two nodes we have in1, out1, in2, out2
18:28 towodo you can look at the intersections of these sets, and maybe invent a formula based on sizes of intersections (and arithmetic such as /)
18:29 towodo considering only those OTUs that occur in both trees I suppose
18:30 kcranstn both trees?
18:31 towodo if an OTU is only in one tree it’s not informative, right?
18:31 towodo if in1 subset of in2 or vice versa then 0 conflict
18:31 towodo so, look at sizes of in1-in2, in2-in1, and in1 intersect in2,  maybe among those there’s a ratio of interest
18:32 towodo maybe there’s a bayesian model of this…
18:32 towodo if in1 and in2 are disjoint then no conflict as well
18:33 towodo I don’t know, maybe the non-shared OTUs are significant
18:34 kcranstn no, I don’t think so
18:42 * towodo goes to vote
18:42 kcranstn jealous
19:09 kcranstn joined #opentreeoflife
19:19 kcranstn joined #opentreeoflife
19:20 mtholder joined #opentreeoflife
19:40 kcranstn joined #opentreeoflife
20:07 towodo joined #opentreeoflife
21:06 kcranstn joined #opentreeoflife
21:20 jimallman joined #opentreeoflife

| Channels | #opentreeoflife index | Today | | Search | Google Search | Plain-Text | summary