Friday, May 27, 2011

first friday

i feel like i spent a lot of thursday spinning my wheels due to various barriers to coding. in the morning, i was working on the script and it stopped working on my office desktop. i ran it once, it worked, i changed the input file folder and it started throwing an error from within the script ("map() requires two arguments", or something like that.) i spent a fruitless... half hour? hour? ... trying to figure out how i had broken the script that one minute prior had been working but then gave up and went home to my laptop (which ran the exact same script beautifully).

after that,  i was beginning to figure out how to "re-access" fasta files once a path is saved in mongoDB, which led me to trying to parse json formatted data, which led me to trying to figure out how to get the module simplejson, which lea to setuptools, which led to frustration and a fundamental lack of understanding. 

that took care of thursday and when i went in to my office this morning, the script was throwing the same error and i'm not real sure what to do about that. i don't understand what's happening that could cause an error originating in the script to show up on one machine and not the other. the office machine had some problems with mongoDB because i accidentally installed it more than once and the poor computer was getting its wires crossed, so it may have something to do with that, but i don't know how. 

again, i came home to my laptop and i have made the most satisfying progress in 48 hours this afternoon. i finally successfully installed setuptools and was thus able to get simplejson. i also installed pysam, for sam/bam files. so i've written two very small scripts that [1] can create sequence objects from a fasta file whose path is in mongoDB and [2] print data from a subset of reads in an indexed bam file that align to a given locus. although they are clumsy and there are probably better ways to achieve both these goals, i feel good about having these rather important initial steps taken. 

OH. and technically the 48 hours of frustration began on wednesday afternoon when i tried to come up with a protocol for committing code to github. what a disaster! i managed to brute force my way through it and gain some understanding of git/github but it wasn't pretty. my account looks very silly, with files that are tagged incorrectly and lead to nothing and i haven't figured out how to correct them yet. i think i'll spend some time this afternoon trying to commit my two new baby scripts and hopefully i can do it quicker and more correctly than wednesday! 

No comments:

Post a Comment