Camelia, the Perl 6 bug

IRC log for #bioperl, 2010-05-13

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

All times shown according to UTC.

Time Nick Message
00:04 abooz damn. I don't get how git on eclipse works...
00:06 deafferret :)   vi works great
00:32 abooz :P vi is nice sorting things on a server from a serial console
00:32 abooz developing i need something more cluttered.
00:32 abooz to distract me ;)
00:42 deafferret ping git gurus
00:43 deafferret hmm... irc.perl.org #moose will be my #git-help I think  :)
00:44 luisrodr joined #bioperl
01:27 luisrodr joined #bioperl
03:24 github joined #bioperl
03:24 github bioperl-live: master Chris Fields * 277039d (9 files in 4 dirs): commit some simple fixes to EUtilities print output - http://bit.ly/9MZy3E
03:24 github left #bioperl
03:24 github joined #bioperl
03:24 github bioperl-live: topic/bug_3077 Chris Fields * 277039d (9 files in 4 dirs): commit some simple fixes to EUtilities print output - http://bit.ly/9MZy3E
03:24 github left #bioperl
03:49 github joined #bioperl
03:49 github bioperl-live: topic/bug_3077 Chris Fields * 13af401 (2 files in 2 dirs): fix coordinate logic error, patch courtesy Ha X. Dang - http://bit.ly/cbaexK
03:49 github left #bioperl
03:53 github joined #bioperl
03:53 github bioperl-live: master Chris Fields * 13af401 (2 files in 2 dirs): fix coordinate logic error, patch courtesy Ha X. Dang - http://bit.ly/cbaexK
03:53 github left #bioperl
04:27 cawiss joined #bioperl
04:28 cawiss left #bioperl
04:31 cawiss joined #bioperl
05:09 luisrodr joined #bioperl
05:59 github joined #bioperl
05:59 github bioperl-live: master Heikki Lehvaslaiho * 2618d3e (0 files in 0 dirs): initial list of file types we do not want in the repository, e.g. temporary files
05:59 github bioperl-live: master Heikki Lehvaslaiho * 4e981dc (1 files in 1 dirs): Reformatting; minor edits; use canonical capitalization for BioPerl
05:59 github bioperl-live: master commits 13af401...4e981dc - http://bit.ly/b6WRsn
05:59 github left #bioperl
07:40 buelsbot joined #bioperl
08:35 vinnana joined #bioperl
11:04 cawiss joined #bioperl
12:00 deafferret .o.
12:00 * deafferret meditates on topic/bug_*
12:43 Jun How can I find the other soc students' information? e.g. their coding blog?
12:45 deafferret dunno. it's not on the GSoC wiki pages?
12:46 deafferret isn't there a mailing list (or three?)
12:49 github joined #bioperl
12:49 github bioperl-live: topic/bug_2515 jhannah * 22756e1 (0 files in 0 dirs): [2515] Trying to get the ball rolling, but already tripping on next_seq(). ... - http://bit.ly/cCpSUw
12:49 github left #bioperl
12:49 deafferret wow. that's a fast IRC bot  :)
12:49 deafferret like.... 1 second lag  :)
12:53 Jun There is a mailing list, but there is no personal information there
12:53 deafferret send an email to that list asking people to send you their info?
12:58 Jun ok, i see
14:04 luisrodr joined #bioperl
15:00 Jun joined #bioperl
15:12 vinnana left #bioperl
15:43 perl_splut joined #bioperl
16:07 spekki01 anyone have any experience using local-lib?
16:07 deafferret just replied to you in irc.perl.org #local-lib  :)
16:08 spekki01 lol thanks
16:13 luisrodr joined #bioperl
16:56 cawiss joined #bioperl
16:58 faceface rbuels: busy?
16:59 rbuels faceface: semi-busy, what's up
17:00 faceface I want to pick your brains for an alg...
17:00 faceface prolly solved
17:00 * rbuels opens the hinge on the top of his cranium
17:00 rbuels well, hinged....hatch?
17:00 rbuels or something.
17:00 rbuels (it's pining for the fjords)
17:00 faceface I have a bunch of sequence anchored markers on a bunch of sequecnes. the markers are ordered in two maps. I want to layout the sequences 'optimally' relative to the two coordinate systems.
17:01 rbuels there, see, now you don't remember my gaffe, you remember that sketch.
17:01 rbuels optimally.
17:01 * faceface remembers to be carefull around rbuels
17:01 faceface yes
17:01 faceface ;-)
17:02 rbuels so, you are the poor guy that has to do the final physical mapping for potato?
17:02 rbuels congratulations, i'm sorry
17:02 rbuels ;-)
17:02 faceface so like the map closest to my org has higher weight, and gets the deciding vote, but suppose one sequence has markers at pos  5 5 5 6 6 7 and another has markers at 6, it should come after
17:02 faceface ;-P
17:03 faceface and if that sequence also has anchors in the second map, ...
17:03 faceface you ken the idea?
17:03 rbuels yeah.
17:03 faceface I heard you do something similar on sgn?
17:03 rbuels not really.
17:03 rbuels well, lukas did something ...
17:03 faceface I figure it's basic network layout optimization
17:03 rbuels probably was off of only one map though
17:03 faceface I was wondering if I should ask in there
17:04 rbuels eh, he's probably not going to pay attention to irc anyway
17:04 rbuels he had a meeting to go to or something
17:04 * faceface dances around nacked
17:04 faceface oh
17:04 * faceface stops dancing
17:04 rbuels o_O
17:04 rbuels lol
17:05 * rbuels experiments writing some pseudocode for this
17:05 faceface There is a whole group of us mapping, but I'd like to solve this particular problem
17:05 faceface I have something reasonable, its just annoyingly 'off'
17:05 * faceface pasts
17:05 rbuels off?
17:05 rbuels the results are off?
17:06 rbuels (and how do you determine they are off?)
17:06 rbuels well really, you are generating a consensus map, because the sequence is reallly just another map
17:07 rbuels for the purposes of this.
17:07 faceface right
17:07 rbuels so i would start by converting all the maps to sequence coordinates.
17:07 faceface I'm using a comparative map viewer to view it all, and I define off as when I say, 'oh that should be lower down'
17:07 rbuels making up some sequence coordinates for the genetic maps ...
17:07 rbuels and then once they are all in the same coordinate system, the fun can begin ...
17:08 faceface http://pastebin.com/d9hyHWvV
17:08 luisrodr joined #bioperl
17:08 rbuels luisrodr: #obf-soc?
17:08 faceface $ss = super-scaffold
17:09 luisrodr rbuels: No, I'm not at the GSoC
17:09 faceface note my normalized rank
17:09 faceface line 59
17:09 faceface NOTE IT!!
17:09 faceface I've been told I'm too childish
17:10 rbuels probably by someone dutch.
17:10 faceface wait... the words were 'immature'
17:10 rbuels a dour lot, they are.
17:10 faceface ;-)
17:10 faceface nah, I'm just kidding
17:10 faceface sorry for the lunacy
17:10 rbuels lunacy is the only way to stay sane.
17:11 * faceface makes a mental note
17:11 rbuels gah.   do not use map for side-effects.
17:11 rbuels map {$avg += $_} @map_pos;  # NO NO
17:11 rbuels $avg += $_ for @map_pos;
17:11 rbuels # yes
17:11 faceface hmm
17:12 faceface the latter option never came to mind
17:12 * rbuels orders faceface to go learn haskell
17:12 faceface how do you strip "'s of of CSV fields?
17:12 faceface oops
17:13 faceface can I say "my $avg += $_ for @map_pos;"?
17:13 faceface I don't grok how my binds
17:13 deafferret tias    SPOILER: no
17:13 faceface hello deafferret
17:13 * deafferret hides
17:13 * faceface seeks
17:14 rbuels faceface: it should be my $avg = 0;  $avg += $_ for @map_pos;
17:14 faceface right, I was just curious
17:15 * rbuels continues looking at the alg
17:16 faceface want to see input?
17:18 rbuels faceface: no i can see the tab-sep cols from the code
17:18 faceface get me!
17:19 faceface yyy didn't really work
17:19 faceface xxx is quite good...
17:19 faceface what I want is minimal crossing lines, which I'm sure is a solved problem in 2d graph layout, letalone 1d
17:20 faceface oh... immagine one map either side of the sequences
17:20 faceface its a 1D search of sequence order to give fewest crossing lines
17:22 rbuels faceface: so your output here is just an ordering of the scaffolds
17:23 rbuels faceface: what about flipping some of the scaffolds?
17:23 rbuels faceface: and what about assembly errors?
17:24 * faceface tries to breath deeply
17:24 faceface we have about 40 assembly errors
17:24 faceface dang... g2g in 2 mins
17:24 faceface ordering... yeah...
17:24 faceface lemmy show you an image
17:26 faceface http://imagebin.ca/view/B-JVzI.html
17:26 rbuels faceface: maybe you could run some kind of pre-built consensus mapping thing on it, and then postprocess that?
17:27 faceface you mean like for assembly reconciliation?
17:27 rbuels faceface: there must be some high-powered consensus mapping stuff somebody has already written
17:27 faceface right
17:27 rbuels actually, i was thinking genetic consensus maps ...
17:27 faceface rbuels: got any nice tomato maps for me today?
17:27 faceface rbuels: IC
17:27 rbuels faceface: well i heard solgenomics.net had a number of nice maps
17:27 rbuels their code is crap though
17:27 rbuels ;-)
17:28 * faceface tries to keep teh green grocer feel going, but fails
17:28 faceface erm... yeah, I'm combining tom rh dm maps
17:28 faceface (EXPEN 2000)
17:28 faceface I'm just wondering if you know how the pseudo molecule construction is being done
17:29 faceface I'm guessing what I have + some manual work is enough for the time being... I'll try combining it with some loverly tomato super-scaffolds
17:29 rbuels i don't ... i think for tomato either WUR (aka KeyGene) is doing it, or some guy in arizona ...
17:30 faceface took me about 2 hours (including development time) to do one chrome
17:30 faceface ahh... key gene
17:30 faceface nice
17:30 faceface night
17:30 rbuels night
17:33 * deafferret rejects that assertion
17:35 rbuels assertion failed.
18:03 dnewkirk joined #bioperl
18:10 luisrodr joined #bioperl
18:42 frogzilla joined #bioperl
19:19 spekki01 when using eutilities-esearch i get the error http://codepad.org/J1IJDBP3 and im not sure why. I start with an array of 839 accession numbers but basically do esearch on subarrays of 101 element each, each of the other blacks of 101 work but for some reason it breaks on submitting the 404 to 504 block.
19:19 spekki01 what can cause that error message is what im wondering?
19:28 perl_splut joined #bioperl
19:29 rbuels spekki01: that sounds like an error on the ncbi size
19:29 rbuels spekki01: try submitting 99 idents at a time maybe?
19:29 rbuels spekki01: isn't there anyplace less annoying you can download this stuff?
19:29 * rbuels hates ncbi
19:30 spekki01 um, im not sure im just dling it from where my boss tells me to get the records from
19:31 spekki01 but ill try reducing the size and see what happens
19:35 rbuels for as much money and manpower as they seem to have, they don't seem to do a great job
19:36 rbuels maybe i'm being too hard on them though
19:36 perl_splut joined #bioperl
19:58 perl_splut joined #bioperl
20:22 abooz joined #bioperl
20:25 * deafferret votes rbuels CTO of NCBI
20:35 dnewkirk rbuels promises to make your wildest dreams come true?
20:40 deafferret again. yes.   :)
21:13 spekki01 so i have a local install of bioperl 1.6.1, installed via the Build.PL. Prior to this i installed local::lib. Which puts anyting i install with cpan into my local perl module directory and i then installed XML::Twig with cpan and it successfully installed in my local perl module directory. Now whenever i run my script i get the error http://codepad.org/5hqsEBkM i have a feeling that Bio::DB::Taxonomy::entrez doesnt know to look locally f
21:13 spekki01 or the twig module. Any thoughts?
21:15 rbuels hmmm
21:16 rbuels if you do 'use XML::Twig' at the top of your script, does it work?
21:16 rbuels spekki01 ^^
21:16 spekki01 lemme give it a try
21:16 spekki01 should i also specify where to look for it?
21:16 rbuels that would indicate whether the thing is actually not installed, or if bioperl is off its rocker in claiming that it's not
21:16 rbuels spekki01: you should just run your script like you have been ..
21:17 spekki01 kk
21:17 perl_splut looks like BioPerl is installed in the system library folder
21:18 perl_splut not sure why you'd need local::lib for that
21:18 rbuels oh ho, didn't notice that.
21:19 spekki01 hmmm it might be an older version still on the server
21:19 rbuels spekki01: is your script using that local::lib?
21:19 rbuels spekki01: or are you doing -Mlocal::lib when you invoke it?
21:19 spekki01 neither
21:19 rbuels spekki01: or are you setting the environment variables according to the local::lib docs?
21:19 spekki01 i set the environment variables
21:20 rbuels echo $PERL5LIB
21:20 rbuels what is in it?
21:20 spekki01 um that cant be right its empty.....
21:20 rbuels uh huh.
21:21 rbuels local::lib just prints env variables, it doesn't set them for you
21:21 rbuels you have to set them
21:21 spekki01 i feel like shooting myself
21:21 perl_splut hmm... Macs default to csh not bash
21:21 rbuels spekki01: if you shoot yourself you will not be able to sit in this channel and help others as pennance.
21:21 perl_splut and the variable doesn't survive the terminal window being closed
21:22 spekki01 sigh, fine ill keep grinding away at this lol
21:23 spekki01 so im gonna assume set it via PERL5LIB=/home/path/lib; export PERL5LIB ?
21:23 rbuels the overly-smart (and less informative) Taxonomy error message isn't helping though
21:23 deafferret rbuels: that depends on what he shoots himself with
21:23 perl_splut that only works if you're using Bash, spekki01
21:23 rbuels spekki01: do perl -Mlocal::lib=path/to/your/local-lib and it will print the env vars for you to set
21:23 spekki01 yeah i beleive i am, now i also get why this had worked earlier before i opened a new terminal window
21:24 rbuels spekki01: and then go read the local::lib docs on search.cpan.org.
21:24 rbuels spekki01: very closely.
21:24 rbuels lol
21:24 deafferret spekki01: is you .bashrc stuff not getting set in new windows?
21:24 rbuels well that would assume he's putting this in this .cshrc or .bashrc ...
21:25 deafferret rbuels: he put stuff in his .bashrc earlier today
21:25 deafferret while you were ignoring him
21:25 deafferret tsk tsk
21:25 * rbuels puts rm -rf ~ in deafferret's .bashrc
21:25 perl_splut if it is in the .bashrc and isn't surviving then probably not using bash
21:25 spekki01 it seems like its set in the .bashrc file
21:26 perl_splut or bash isn't reading .bashrc and is instead reading some other file, like .profile
21:26 deafferret spekki01: open a new window.    echo $PERL5LIB
21:26 rbuels spekki01: which tests if it's getting set in your .bashrc
21:26 spekki01 kk one sec
21:27 spekki01 that echo gives me nothing
21:27 spekki01 but when i cat .bashrc i get http://codepad.org/9vzUKSET
21:27 perl_splut echo $SHELL
21:27 spekki01 /bin/bash
21:28 rbuels egads.  mac servers.
21:28 spekki01 yeah they are
21:28 * rbuels says 3 hail marys
21:29 perl_splut so, it isn't reading .bashrc
21:29 rbuels i dunno, that looks ok ...
21:29 rbuels yeah that would be my guess.
21:29 perl_splut so it must be reading .bash_profile
21:29 rbuels oh could be.
21:29 perl_splut man bash
21:29 perl_splut read through that to figure out when/why it reads one vs the other
21:29 rbuels like ... a feminist?
21:30 spekki01 .... weird i just did source .bashrc and then that echo $PERL5LIB and it spit out the variables that it should have
21:30 perl_splut more like a program that Apple had to mess with
21:30 rbuels probably
21:30 rbuels you could just symlink your .bashrc and .bash_profile
21:30 perl_splut sourcing the file causes it to be read
21:30 spekki01 im not sure where that .bash_profile is?
21:31 rbuels spekki01: in your home dir, should be
21:31 perl_splut if it doesn't exist then it isn't being read, heheh
21:31 rbuels if it's not there, ... yeah
21:31 perl_splut don't think one exists by default
21:31 spekki01 not there :(
21:31 rbuels you could just ln -s ~/.bashrc ~/.bash_profile
21:33 spekki01 that symlink looks like it did the trick
21:33 spekki01 its getting loaded now when i open a new terminal window
21:33 spekki01 thanks again guys
21:34 perl_splut might want to pick up the Animal book on Mac OS
21:35 perl_splut http://oreilly.com/catalog/9780596153281/
21:42 spekki01 thanks
21:57 rbuels deafferret: yapc10hackathon branch has all been merged in, i'm going to blow it away
21:57 * rbuels takes aim
21:58 deafferret rbuels++
21:58 * rbuels waits for deafferret to say 'noo!'
21:58 * rbuels fires
21:58 deafferret noo!
21:58 deafferret I mean,... yay!
21:58 rbuels lol
22:00 deafferret what was in there anyway?  biab   & feed dogs
22:01 rbuels deafferret: just miscellaneous silly bugfixes
22:02 dukeleto joined #bioperl
22:04 rbuels i'm going to move branches that haven't been modified since 2008 to attic/
22:11 deafferret lol
22:11 deafferret I thought I was alone in my righteous anger
22:11 deafferret note to self: next time, just do it
22:13 deafferret i tried lots of communication and feel like the result was "don't overcomplicate things"
22:13 deafferret lesson learned
22:15 deafferret @less_communication is now 'wife', 'bioperl'
22:15 deafferret :)
22:29 rbuels actually i changed my strategy and emailed bioperl-l
22:29 rbuels and i'll give it a day or two
22:29 deafferret pretty weak righteous anger, that   :)
22:30 deafferret do try to man up a little
22:30 rbuels one of these days, alice...
22:31 rbuels *pow*, right to the moon!
22:31 deafferret cheese!
22:31 * deafferret sneaks more testosterone into rbuels' coffee
23:06 kyanardag_ joined #bioperl
23:10 faceface something is in my caffine!
23:10 faceface (can't sleep)
23:13 faceface anyone actually use DBI?
23:13 faceface I stopped years ago
23:15 deafferret :)   ya, I use the heck out of DBI
23:16 deafferret rbuels: you clean your desk?
23:16 rbuels well, sometimes.
23:16 rbuels i damn sure do it when i move.
23:17 faceface deafferret: how do you organize your SQL inside your perl?
23:18 deafferret well, for CRUD I use DBIx::Class
23:18 deafferret when things gets too screwy for SQL::Abstract in my tiny brain I just  <<EOT it
23:49 faceface ty
23:55 faceface I never learned dbix::class
23:55 deafferret there's still time!
23:55 rbuels faceface: no there's not.
23:55 rbuels you're done.
23:56 was kicked by deafferret: rbuels
23:56 rbuels joined #bioperl
23:56 rbuels out to pasture!
23:56 deafferret oops. faceface defense network activated

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