These are chat archives for dereneaton/ipyrad

14th
Sep 2016
Deren Eaton
@dereneaton
Sep 14 2016 16:05
Hey @SheaML , looks like a bug. We've been working on step6 to try to squash these. Can you provide more details. What version of ipyrad are you using, is your data single or paired, at which stage of part6 did it bail out? We should be able to fix it quickly.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:20
Hi @isaacovercast , regarding #184 I have untared and gzipped the file back. The error is still there:
TypeError(expected a string or other character buffer object)
I'll try with the uncompressed file
Deren Eaton
@dereneaton
Sep 14 2016 16:22
@StuntsPT Have you tried running the example data sets? It takes only about 2 minutes and would give us an idea of whether your installation is the problem.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:22
Yes, it's the same with the uncompressed file
@dereneaton will do that now
Deren Eaton
@dereneaton
Sep 14 2016 16:24
@isaacovercast There may be a bug in the zcat_make_temps func, it's working on Master now. But I made some changes to it since 0.3.41. I'll look into it.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:26
it's working in the example data...
let me try relative paths...
Deren Eaton
@dereneaton
Sep 14 2016 16:27
That makes me think it might be a bug in the func above. Its only used on data sets that are big, and so not checked in the sims
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:28
I'll try with the first 10k lines of my file
Deren Eaton
@dereneaton
Sep 14 2016 16:28
ok
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:30
nope, still erroring out
this exception
I can't get a full traceback
but it seems to be coming from "assembly.py"
there is a "try" statement there
among other things it's checking the steps
let me try something else...
Deren Eaton
@dereneaton
Sep 14 2016 16:32
@StuntsPT are you on Master branch?
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:33
no, I am on 0.3.41
any easy way to get master from conda?
or should I just do a very dirty hack, and manually overwrite the installation? Just to test this.
Deren Eaton
@dereneaton
Sep 14 2016 16:35
if you install ipyrad once with conda then you will have all the dependencies. Then you can git clone the repo, cd into it, and run pip install -e . (using the conda version of pip -- which you can check with which pip), and this will tell conda to use the git branch of ipyrad. Then you can hack on it, fix bugs, etc.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:37
just out of curiosity, where will the pip from conda install ipyrad? will it overwrite 0.3.41?
or will it go somewhere else?
Deren Eaton
@dereneaton
Sep 14 2016 16:38
hmm, that's a good question, let me look.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:40
the packages in the conda "site-install" don't seem to be versioned
I'd say 0.3.41 will get overwritten
Deren Eaton
@dereneaton
Sep 14 2016 16:40
the ipyrad executable is in /home/deren/miniconda2/bin/ipyrad.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:41
yeah, it's the wrapper script
Deren Eaton
@dereneaton
Sep 14 2016 16:41
there is an arg called conda develop but I couldn't quite get it to do what I wanted to, at least last time I tried several months ago, and so I settled on this pip strategy to be able to work on a development version. it's been working, but maybe there is a better way.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:42
~/Software/miniconda2/bin/pip install -e git://github.com/dereneaton/ipyrad.git
~/Software/miniconda2/bin/pip install -e https://github.com/dereneaton/ipyrad.git
what am I missing?
I'll take a look at conda develop tonight
but this is my first time using conda
Deren Eaton
@dereneaton
Sep 14 2016 16:43
The period at the end.
to tell it install from (here)
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:44
I still have to get acquainted
oh
right
missed the ".", sorry
Deren Eaton
@dereneaton
Sep 14 2016 16:44
@StuntsPT Thanks for pushing docs updates by the way, super helpful!!
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:45
@dereneaton You're welcome. I hope it helps the project
I used pyrad heavily for the paper I'm currently writing
so thank you!
Deren Eaton
@dereneaton
Sep 14 2016 16:46
glad to hear it
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:46
I'l like to dive into ipyrad's code too
It's quite an awesome project
Deren Eaton
@dereneaton
Sep 14 2016 16:47
Well, it's much much cleaner than pyrad's code, so hopefully it will be easier.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:47
Yeah, I just have less time right now. =-)
ok, so master branch is installed
let's see if the error presists
Successfully installed ipyrad-0.3.41 networkx-1.11
networkx is a new dependency, right?
Deren Eaton
@dereneaton
Sep 14 2016 16:48
yeah
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:48
so I guess I'm installing from master correctlly
the version number remains the same, though
Deren Eaton
@dereneaton
Sep 14 2016 16:49
looksl ike it
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:49
but I guess that gets picked up from the tags
Deren Eaton
@dereneaton
Sep 14 2016 16:49
yeah, it updates only when we push to anaconda.org.
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:49
oh, ok
let's try again
Deren Eaton
@dereneaton
Sep 14 2016 16:49
we have it setup that way through the versioner script
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:50
same error... =-(
Deren Eaton
@dereneaton
Sep 14 2016 16:50
dang...
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:50
would you like my 10k lines fastq that is not working?
Deren Eaton
@dereneaton
Sep 14 2016 16:50
yeah
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:50
it worked with pyrad
Deren Eaton
@dereneaton
Sep 14 2016 16:51
you can email it: deren.eaton@yale.edu
Francisco Pina-Martins
@StuntsPT
Sep 14 2016 16:52
ok
Ok, time to go. I'll take another poke at this issue tonight.
Shea Lambert
@SheaML
Sep 14 2016 19:35
Hi @dereneaton, I was using 0.3.41, on paired-end data, and the analyses seem to fail on the "building database" step (which is also taking many hours).