I hereby claim:
- I am grero on github.
- I am rherikstad (https://keybase.io/rherikstad) on keybase.
- I have a public key whose fingerprint is D07C 7616 0455 F768 B31D 0C91 3885 9141 3121 4871
To claim this, I am signing this object:
Count Overhead File Line Function | |
===== ======== ==== ==== ======== | |
2 0 @Base/boot.jl 385 eval | |
2 0 @Base/client.jl |
import sys | |
import pylink | |
from pylinkwrapper import psychocal, connector | |
from psychopy import sound, visual,event, tools, data | |
from psychopy.tools import monitorunittools | |
import utils | |
import os | |
red = (1.0, 0.0, 0.0) | |
green = (0.0, 1.0, 0.0) |
01/06/15 12:47:18 (pid:3536) About to exec D:\condor\execute\dir_3536\condor_exec.py | |
01/06/15 12:47:18 (pid:3536) Running job as user condor-slot1 | |
01/06/15 12:47:18 (pid:3536) ERROR: failed to produce Win32 argument string from CreateProcess: | |
01/06/15 12:47:18 (pid:3536) Create_Process(D:\condor\execute\dir_3536\condor_exec.py,, ...) failed: | |
01/06/15 12:47:18 (pid:3536) Failed to start job, exiting | |
01/06/15 12:47:18 (pid:3536) ShutdownFast all jobs. |
I hereby claim:
To claim this, I am signing this object:
Process: julia [1558] | |
Path: /Applications/Julia-0.3.0-prerelease-a3248be2eb.app/Contents/Resources/julia/bin/julia | |
Identifier: julia | |
Version: ??? | |
Code Type: X86-64 (Native) | |
Parent Process: bash [1498] | |
Responsible: iTerm [1413] | |
User ID: 501 | |
Date/Time: 2014-04-26 09:49:15.740 +0800 |
Using git-annex to keep track of my movie files. i'm just getting to know the basic workflow, but basically, the actual files are located under .git/annex/objects, while the filenames symlink to these. If files go bad, they end up in .git/annex/bad. To synchronize between remotes (without copying the actual files; this just lets annex know which files are, or should be, available), just use the normal git fetch/merge/commit workflow |
Fisher and shannon information in finite neural populations. | |
Stuart Yarrow, Edward Challis, and Peggy Seriès. | |
The precision of the neural code is commonly investigated using two | |
families of statistical measures: Shannon mutual information and | |
derived quantities when investigating very small populations of | |
neurons and Fisher information when studying large populations. These | |
statistical tools are no longer the preserve of theorists and are | |
being applied by experimental research groups in the analysis of |
#use this command to submit a bunch of jobs decoding a tetrode data set, in this case P442g0002. {1..35} iterates over the number of highpass files. A wrapper job is submitted that waits for all the compute jobs to finish, then issues an email notifying me that the jobs finishes | |
echo "P442g0002 done\!" | qsub -j y -V -N hmmDecodeP442g2 -o $HOME/tmp -m e -M roger.herikstad@gmail.com -hold_jid `for i in {1..35}; do echo "cd $PWD;/opt/cluster/tmp/run_hmm_decode.sh /Applications/MATLAB_R2010a.app/ hmmsort/P442g0002 10000 1e-20 SourceFile highpass/P442_highpass.$( printf %.4d $i ) DescriptorFile P442_descriptor.txt hdf5 save Group 2" | qsub -j y -V -N hmmdecodeP442g2_${i} -o $HOME/tmp/; done | awk '{print $3}' | paste -s -d "," -` | |
#same as above except loops through groups and checks for already computed data | |
for g in {1..4}; do echo "P443g000$g done\!" | qsub -j y -V -N hmmDecodeP443g$g -o $HOME/tmp -m e -M roger.herikstad@gmail.com -hold_jid `for i in {1..605}; do test -e hmmsort/P443_g$( printf %.4d $g ).$( pr |
This is a test. I can't really be sure what I'm going to write, so I'll just stop here. Well, maybe I'll write just a little more. I'm not sure if this works, though, having some issues with syncing here. Ok, apparetly it does, so I guess I'll stop this rubbish and get on with something sensible. |