Perl 6 - the future is here, just unevenly distributed

IRC log for #opentreeoflife, 2014-08-15

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

All times shown according to UTC.

Time Nick Message
00:51 josephwb joined #opentreeoflife
00:55 josephwb i wonder if we should change our standard to 1.9.8, rather than 1.9.5, since the latter is so difficult to obtain
01:19 jimallman It is odd that neo4j doesn’t make it easier to find historic releases. How hard is it to install from GitHub (https://github.com/neo4j/neo4j/releases/tag/1.9.5))?
01:35 towodo joined #opentreeoflife
01:45 josephwb joined #opentreeoflife
01:59 josephwb jimallman: any updates
02:00 josephwb ?
02:00 jimallman not since my email ~9:30pm
02:00 jimallman it’s all good
02:00 josephwb * checking email *
02:00 jimallman i clobbered all the db files and re-extracted them from the tgz file. success!
02:01 jimallman something must have gotten corrupted in the first attempt at untar’ing (ran out of disk space)
02:01 josephwb wee!
02:02 josephwb i was not convinced about the possible version issue
02:02 josephwb but it is good that it came up
02:02 jimallman yeah, it’s good now. sorry for the unneccessary detective work, but it was interesting to discover the version differences.
02:02 jimallman the more you know...
02:02 josephwb yes!
02:03 josephwb anyway, in case you didn't look at my email, i got a copy of 1.9.5 from download of ot10, so i will use that from now on
02:03 jimallman sadly, the new db removed my handy example of a missing taxon, since both Drosophila taxa are now in synthesis.
02:03 josephwb oh, i can get you an example
02:04 jimallman ah, i see. and as jar pointed out, the url for its tarball is in the install-neo4j.sh script
02:04 jimallman cool, thanks for any examples
02:04 josephwb the new tree is displayed on devtree?
02:05 jimallman yep
02:05 josephwb example: "Passeroidea"
02:05 jimallman oh, i need to re-index oti…
02:06 josephwb it is a higher avian taxon that is not supported by the input trees
02:06 josephwb http://devtree.opentreeoflife.org/opentree/argus/ottol@176458/Passeroidea
02:06 jimallman hm, it’s showing up in the synth-tree viewer, but isolated from other nodes. i have one of those already, but i no longer have one that’s missing from synthesis entirely
02:06 jimallman (= deprecated?)
02:07 josephwb "missing from synthesis entirely"?
02:08 jimallman that’s my (shaky) understanding, yes. like this one on production: http://tree.opentreeoflife.org/opentree/argus/ottol@5554385/Drosophila
02:08 josephwb if a taxon cannot be found, it must have been filtered out
02:08 josephwb hmm
02:09 jimallman right. i explain these away (in the new UI) by saying “This is probably incertae sedis, but here are the flags assigned for this taxon and its taxonomic ancestors”
02:09 josephwb so, i grepped ott2.6 (that the prod tree was built from), and I get no hit
02:09 josephwb grep 5554385 taxonomy.tsv
02:09 jimallman the flags are displayed similarly to what you can see in the property inspector on this page: http://devtree.opentreeoflife.org/opentree/argus/ottol@176458/Passeroidea
02:10 josephwb but it is there for ott2.8draft5
02:11 jimallman ah, sounds like i’ve got it wrong. it’s not missing from prod synthesis, it was dropped from the merged taxonomy on prod (ott2.6)
02:11 josephwb towodo could probably easily get a list of new taxa to 2.8 to compare against
02:12 josephwb yeah, i can't find it in ott2.6
02:12 jimallman right, but i need something that is missing from ottt2.8draft5 (newest merged taxonomy)
02:12 josephwb not flagged: not there at all
02:12 jimallman hmmmm
02:13 josephwb something that is filtered out?
02:13 jimallman then how does it turn up on taxon search?
02:13 josephwb a "bad" taxon?
02:13 josephwb hodl on
02:13 josephwb hold
02:13 josephwb on
02:13 josephwb it comes up because i imagine you are querying the new taxonomy
02:14 josephwb ?
02:14 josephwb what are you querying?
02:14 josephwb taxomachine?
02:14 josephwb the treemachine DB?
02:14 jimallman taxon search == TNRS == taxomachine
02:14 josephwb ok
02:15 jimallman specifically, http://devapi.opentreeoflife.org/taxomachine/v1/autocompleteBoxQuery
02:15 josephwb a couple of possibilities
02:15 jimallman oti probably needs an update, but that’s unrelated (afaik)
02:15 josephwb 1) taxomachine ott version was newer than the treemachine DB
02:15 josephwb 2) taxomachine is not filtering like treemachine
02:15 jimallman oh! i remember reading that TNRS includes deprecated taxa. any chance these are filtered out of the taxonomy.tsv file?
02:15 josephwb we probably want no. 2
02:16 josephwb yes
02:16 josephwb yes to filtering
02:16 jimallman 1) is interesting, though...
02:16 josephwb i think this was the case with Drosophila
02:17 josephwb i can confirm it was not filtered. it simply was not there.
02:17 josephwb must be new to ott2.8draft5
02:17 josephwb again, towodo could confirm
02:18 josephwb to be clear: i think 1) explains the Drosophila example, not filtering
02:18 josephwb so, anyway: you want an example of something that is *not* in the new DB
02:18 josephwb right?
02:18 josephwb not just an orphaned node
02:19 jimallman correct. it should not show an isolated node. it should show as a missing tree in the viewer, a la http://tree.opentreeoflife.org/opentree/argus/ottol@5554385/Drosophila
02:19 josephwb okay, we can get examples through looking at flags in the taxonomy.tsv file
02:19 jimallman (this is not super-important, just need it to show the new feedback w/ taxon flags and lineage, for review)
02:20 josephwb anything "barren" won't be in the DB
02:20 josephwb (most things) "extinct", likewise
02:20 josephwb also, "not_otu"
02:20 josephwb anyway, easy
02:20 josephwb just a sec
02:20 jimallman ah, cool. i’m not clear enough on the rules. is this in github somewhere?
02:21 jimallman oh, i’m a dope. it’s on files.opentreeoflife.org!
02:21 josephwb *always the last place you'd think*
02:22 jimallman yea, kinda
02:22 jimallman oh wait, i don’t see ott2.8 here!  http://files.opentreeoflife.org/ott/
02:22 jimallman newest is 2.7. i am redeemed.
02:23 josephwb hmm. we may be using the "Experimental version"
02:24 josephwb *checking*
02:25 josephwb yes: confirmed. we are using "experimental"
02:25 josephwb i think 2.7 had tons of problems, was short-lived
02:25 jimallman ah 2.7, we hardly knew ye
02:26 josephwb verily
02:27 josephwb we should get towodo to update that
02:28 josephwb since you got things working (thanks!), i guess i do not need to re-run things through the newer neo4j immediately
02:28 jimallman not a crisis, no
02:29 josephwb good
02:29 jimallman i’m grabbing the experimental 2.8 tgz, will dig for my own examples (thanks for the tips on flags that matter)
02:30 josephwb i looked for some barren taxa, but they were not found
02:30 josephwb perhaps taxomachine is doing *some* filtering
02:30 josephwb i.e. "barren" == "bad"
02:31 jimallman et voila!  http://devtree.opentreeoflife.org/opentree/argus/ottol@4675814/Caryocarididae
02:33 josephwb not clear why some things i tried didn't work
02:33 josephwb oy
02:36 jimallman yeah, i’m hazy on the rules for deprecated taxa. but EXTINCT_DIRECT seems to do the trick!
02:36 josephwb great. good to know.
02:37 josephwb (i am also hazy on the flags)
02:38 jimallman at least i’m assuming that EXTINCT_DIRECT was the trump card. Carycarididae is flagged (sibling_lower,extinct_direct,extinct_inherited)
02:38 jimallman oh, quickest path might be to just pick from hidden.tsv, right?
02:38 jimallman (i just noticed this taxon is listed in both files)
02:38 josephwb of course. doh.
02:38 josephwb never looked at that myself
02:39 jimallman or deprecated.tsv  :-/
02:39 jimallman it’s official, i think i’m done for tonight.
02:39 josephwb i can take a look at the treemachine code and tell you for certain which taxa are dropped
02:39 josephwb me too!
02:40 jimallman ok, but it’s clear that some taxa are blocked/deprecated/hidden/whatever in taxomachine, too
02:40 josephwb good night jimallman. and of course ilbot3!
02:40 josephwb yes, taxomachine is hazy
02:40 jimallman but yea, it would be good to know the Flags That Really Matter, if we want friendlier messages in the UI.
02:40 josephwb verily
02:41 jimallman thanks again for your time today. see you tomorrow!
02:42 josephwb no, thank you. tomorrow, then
02:49 josephwb joined #opentreeoflife
12:23 josephwb joined #opentreeoflife
12:39 josephwb joined #opentreeoflife
12:40 josephwb joined #opentreeoflife
13:00 towodo joined #opentreeoflife
13:26 kcranstn joined #opentreeoflife
14:12 josephwb joined #opentreeoflife
16:47 towodo joined #opentreeoflife
16:51 josephwb would be nice to be able to specify a custom context for name mapping in the curator
16:52 josephwb i.e. send it to taxomachine, it is smart enough to subset some existing context
16:54 josephwb thinking of paraphyletic taxa, like "fish"
17:22 jimallman josephwb: As i recall, the search contexts need to be a predetermined list.
17:26 josephwb right, that is how it is set up. i am thinking taxomachine could be made smarter
17:26 josephwb (the contexts were built to save time i.e. don't search everything)
17:26 josephwb working within an existing context should be feasible (i think)
17:27 josephwb just thinking out loud
17:28 josephwb was working with a shark study, and the closest existing thing ws "vertebrates"
17:28 jimallman yeah, that’s a pretty broad brush
17:29 towodo there will always be groups of interest that aren’t well corraled by any predefined set of groups.
17:29 josephwb i didn't run into any obvious problems, but not being familiar with sharks, i was worried
17:29 towodo if it’s not sharks, it’s hagfish
17:29 josephwb right. my thought was to send "sharks" as a custom context to taxomachine
17:30 towodo oh.
17:30 josephwb it will find vertebrates as the closest context, and work within that
17:30 towodo by taxon name?
17:30 josephwb yes.
17:30 josephwb "Chondrichthyes"
17:30 josephwb something in ott, obviously
17:31 josephwb perhaps a dropdown link on the treeviewer; can only match names in ott
17:31 josephwb list, not link
17:32 josephwb anyway, just a thought
17:32 josephwb i would have to talk to cody to see how difficult it might be
17:34 josephwb above should read: perhaps a searchbox like the one on the treeviewer
17:41 kcranstn joined #opentreeoflife
18:11 kcranstn joined #opentreeoflife
20:45 towodo joined #opentreeoflife
22:57 josephwb joined #opentreeoflife
23:33 josephwb joined #opentreeoflife

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