Open gwaybio opened 6 years ago
We will test each module to see if the packages are available.
linked with #146 #147
The kernel workflow works! Please add packages (python, R, command line, etc.) and specific versions here
Also related to #122
Also related to #77
sklearn - i think that some of the ML exercises will break with the move to any recent version (that's ok). I had convert them last year to an older version b/c sagemath was running something quite old
I'd go with the current release (0.19) and we'll fix the exercises.
added HOMER 4.9.1
https://bioconda.github.io/recipes/phylip/README.html
:point_up: this is Phylip
Just to make sure - are we sure we know how to distribute this to students as well? If there instructions they'll need to run, then we need to put them in the first exercise.
they're running everything in cocalc correct, not on their local machines? It should be automatically distributed in the notebooks. Command line interfaces outside of notebooks are trickier but we can do 1 of 2 things.
e.g.
alias samtools='/home/user/.conda/envs/python-kernel-GCB535/bin/samtools'
# in .bashrc
alias setenv='anaconda3 && source activate python-kernel-GCB535'
i guess I should confirm this a the student account. on it!
I think that if you are asking students do so this every time then it is unlikely to occur. Things need to be as easy as opening a terminal (so we may need to add it to their bashrc files somehow). There may be a way to initially define a bashrc file that they inherit from.
bedtools
ugh... the kernels are NOT automatically transferred to the student users.
Unknown kernel python-kernel-GCB535 (select a valid kernel from the Kernel menu) FileEditViewInsertCellKernelHelp
Either we use the previous solution with built-in kernels and request packages when we need them or I dig deeper into figuring out how to transfer custom kernels to students (wasn't there a technical contact from cocalc for these kind of things?)
Click the help button in the upper right to start a ticket. They're quick.
On Thu, Dec 21, 2017 at 3:50 PM Greg Way notifications@github.com wrote:
ugh... the kernels are NOT automatically transferred to the student users.
Unknown kernel python-kernel-GCB535 (select a valid kernel from the Kernel menu) FileEditViewInsertCellKernelHelp
Either we use the previous solution with built-in kernels and request packages when we need them or I dig deeper into figuring out how to transfer custom kernels to students (wasn't there a technical contact from cocalc for these kind of things?)
— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/greenelab/GCB535/issues/155#issuecomment-353454590, or mute the thread https://github.com/notifications/unsubscribe-auth/AAhHs6y01TX4fEa6Exc4Q82Qybu_PyQkks5tCsSbgaJpZM4RCgAt .
-- Casey S. Greene, Ph.D.
Assistant Professor Dept. of Systems Pharmacology and Translational Therapeutics Perelman School of Medicine University of Pennsylvania web: http://www.greenelab.com phone: 215-573-2991
Director Childhood Cancer Data Lab Alex's Lemonade Stand Foundation web: http://ccdatalab.org
Ticket submitted (ticket num: 4606)
- We are interested in creating custom jupyter kernels with conda environments. We have successfully created a custom Jupyter kernel using these instructions https://github.com/sagemathinc/cocalc/wiki/Programming#anaconda-jupyter.
- Our goal is to perpetuate this environment to students, but it seems this is not done automatically.
- Is there functionality to define a custom kernel, send to students, and be able to update between assignments? We have discussed this issue in some amount of detail here: https://github.com/greenelab/GCB535/issues/155
I moved this to the 2019 milestone, but we should add to this list throughout the 2018 semester
We need to confirm that all of the packages and dependencies exist in the cocalc environment
We should explore specifying a custom conda environment https://github.com/sagemathinc/cocalc/wiki/Programming#anaconda-jupyter
Packages