search

berkeley-dsep-infra / datahub

JupyterHubs for use by Berkeley enrolled students
https://docs.datahub.berkeley.edu
BSD 3-Clause "New" or "Revised" License
65 stars 39 forks source link

Request to have two packages installed on the dataHub #653

Open mdshapiroLBL opened 5 years ago

mdshapiroLBL commented 5 years ago

I would like to request two packages be added to the data hub: ROOT and pythia8. These packages are both widely used by particle physicists and have well tested python interfaces that function well through Jupyter. We plan to use these packages for Jupyter based problems sets in two Physics classes (Physics 129 and Physics 226).

Both packages are available for installation via conda. The relevant information is available at the sites listed below:

ROOT: https://github.com/conda-forge/root-feedstock

pythia8: https://github.com/conda-forge/pythia8-feedstock

These packages are also available in docker containers, but if the conda installation works, that would be a preferable solution.

Please let me know if you have any questions about these packages or any issues with installation.

Thanks, Prof Marjorie Shapiro Physics Department

yuvipanda commented 5 years ago

Thanks for making the requrest here, @mdshapiroLBL!

Do you know if these packages are also available from pypi.org / installable via pip? On datahub.berkeley.org we use pip right now, rather than conda...

Also can you tell us what the timeline for the classes is?

Thanks.

mdshapiroLBL commented 5 years ago

Thanks for the rapid response. We are targeting Fall 2019 for the classes.

I don't think that we can use pip to install. Both ROOT and pythia8 are written in C++ with python bindings. I didn't think pip could handle that (although I would be happy to be told that I am wrong). Conda uses CMake to handle the various OS dependencies.

BTW, I tried to see what version of gcc is native for the dataHub and was surprised to find that when I do:

import os os.system("gcc -v")

From my dataHub account I get:

ModuleNotFoundError: No module named 'gcc'

I have never seen a Linux system without gcc as part of the environment, so I am a bit confused.

ROOT is the more important of the two packages for us, so let's start with that one. If you think using Docker is an easier way to access this package, here is a pointer to the relevant documentation for ROOT: https://hub.docker.com/r/rootproject/root-ubuntu16/

On Sat, Apr 27, 2019 at 3:44 PM Yuvi Panda notifications@github.com wrote:

Thanks for making the requrest here, @mdshapiroLBL https://github.com/mdshapiroLBL!

Do you know if these packages are also available from pypi.org / installable via pip? On datahub.berkeley.org we use pip right now, rather than conda...

Also can you tell us what the timeline for the classes is?

Thanks.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/berkeley-dsep-infra/datahub/issues/653#issuecomment-487325303, or mute the thread https://github.com/notifications/unsubscribe-auth/ALLWGUU3QR5MM2ZDV2JBNXDPSTJLJANCNFSM4HI5A2KA .

ryanlovett commented 5 years ago

CERN provides some binaries for ROOT at https://root.cern/content/release-61206 -- latest supported Ubuntu is 17.x.

yuvipanda commented 5 years ago

@mdshapiroLBL when I run 'gcc -v' in a terminal ('New -> Terminal' from the Jupyter Notebook directory listing page), it tells me we've gcc 7.3 installed. Not sure why os.system isn't finding it.

I think we might be switching to conda anyway in the fall, so it should be ok. If not, we should be able to spin up a new image just for you.

is it ok if we start working on this once this semester ends?

mdshapiroLBL commented 5 years ago

Strange about gcc. That version is fine. Yes, it is OK to wait until after the end of semester to look at this. Please contact me if you have any questions.

On Tue, Apr 30, 2019, 11:37 AM Yuvi Panda notifications@github.com wrote:

@mdshapiroLBL https://github.com/mdshapiroLBL when I run 'gcc -v' in a terminal ('New -> Terminal' from the Jupyter Notebook directory listing page), it tells me we've gcc 7.3 installed. Not sure why os.system isn't finding it.

I think we might be switching to conda anyway in the fall, so it should be ok. If not, we should be able to spin up a new image just for you.

is it ok if we start working on this once this semester ends?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/berkeley-dsep-infra/datahub/issues/653#issuecomment-488066049, or mute the thread https://github.com/notifications/unsubscribe-auth/ALLWGUV3I2NRYFJ67NWPDUTPTCGVBANCNFSM4HI5A2KA .

mdshapiroLBL commented 5 years ago

I wanted to come back to this request. In April we agreed to postpone it until the summer since there were plans to support conda installations then. We'd like to be able to use Root and Pythia8 modules from the datahub for two Physics classes in Fall 2019 and we need some time to test that the notebook solutions properly work before the start of semester.

mdshapiroLBL commented 5 years ago

@yuvipanda @ryanlovett Just wanted to ping you on this ticket.

Cheers Marjorie

mdshapiroLBL commented 4 years ago

This issue has been dormant for a long time, but I'd like to revive it. Several students in my class (Physics 129 which is a data science domain emphasis offering) have requested the ability to use the Root analysis package from the dataHub. As noted earlier in this thread, installation is usually done using conda. At the time we last exchanged messages, you indicated that support for conda installation was expected in the future.

Is it possible to install these packages?

Thanks Marjorie

yuvipanda commented 4 years ago

So sorry for dropping the ball on this one, @mdshapiroLBL!

@felder do you think you have time to look at this? Now that we're using conda for datahub, we should be able to install it from https://anaconda.org/conda-forge/root/.

felder commented 4 years ago

@yuvipanda I tried all day to get the datahub image to build and I simply cannot.

Please see https://github.com/berkeley-dsep-infra/datahub/pull/1811

yuvipanda commented 3 years ago

@mdshapiroLBL are you still interested in using root? If so, I'm tempted to make a separate hub for physics that has it. I think it might complicate our main image a bit too much...

mdshapiroLBL commented 3 years ago

There are only 2 classes where root and Pythia8 would be used (physics 129 and 226) but for those, it would be great! I agree that. Separate hub makes sense. I think they do the same thing at NERSC. For my class, Pythia8 and Root would be great additions for the students' final projects so it would be fine if the hub isn't available at the star of semester.

On Tue, Aug 10, 2021, 10:50 PM Yuvi Panda @.***> wrote:

@mdshapiroLBL https://github.com/mdshapiroLBL are you still interested in using root? If so, I'm tempted to make a separate hub for physics that has it. I think it might complicate our main image a bit too much...

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/berkeley-dsep-infra/datahub/issues/653#issuecomment-896521073, or unsubscribe https://github.com/notifications/unsubscribe-auth/ALLWGUXVOTECDWDE6UNKLH3T4IFTPANCNFSM4HI5A2KA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email .

spagangriso commented 3 years ago

Hi @yuvipanda , I'll be teaching physics 226 this fall semester and would be indeed great for the students to be able to use an image having ROOT and python3 (and pythia8 if possible, but that's lower in priority). Thanks! Simone Pagan Griso