Re: [jupyter] Not able to find R in the Notebook in jupyterhub

JupyterHub does not install additional kernels. If you want R notebooks, you must install the R kernel. With conda, this would be: conda install -c r r-irkernel -MinRK ​ On Mon, Jul 11, 2016 at 1:34 AM, Bond wrote: > Hi > > I'm not able to find R in the Notebook in jupyterhu

Re: [jupyter] change the R version in Jupyter

Jupyter uses the available version of R on your system. However you installed R the first time (apt, conda, etc.) you can investigate installing additional versions of R. Once installed and the `irkernel` package installed with that R, it will be available to Jupyter. -MinRK On Mon, Jul 11, 2016

Re: [jupyter] Looking for a scheduling solutions for a jupyter notebook

You can run notebooks offline with jupyter nbconvert --execute notebook.ipynb [--output destination.ipynb]. To run it every Monday, you could add this command to a cron job. -MinRK On Tue, Jul 12, 2016 at 1:23 PM, Paul A wrote: Are there open source solutions similar to domino that would e.g

Re: [jupyter] Installing JupyterHub from Source

conda and pip can both be used without a network connection. You can make your own repository (directory) of conda packages and/or wheels, and then install from there instead of the internet. For instance, with pip and npmbox: pip wheel jupyterhub npmbox configurable-http-proxy -MinRK

Re: [jupyter] Our licenses have drifted from "pure" BSD, we should fix that...

has never been an entity. But if using the plural "Contributors" text is clearer than the collective "Team", that's fine, too, and changes no meaning. -MinRK On Sat, Jul 23, 2016 at 3:44 AM, Fernando Perez wrote: > Hi all, > > I recently noticed that there'

Re: [jupyter] Installing JupyterHub from Source

http-proxy.npmbox This assumes that you have done a one-time approved install of pip and npmbox on the server. -MinRK > > On Monday, July 25, 2016 at 4:40:23 AM UTC-4, Min RK wrote: >> >> conda and pip can both be used without a network connection. You can make >> your

Re: [jupyter] Multiple Node Environment for Jupyterhub

If you are already using Docker, Docker Swarm may be a good choice, which does its own distributing of containers across machines. You can generally use DockerSpawner unmodified with Swarm. -MinRK On Tue, Jul 26, 2016 at 4:05 PM, Sudipto Goswami wrote: > I am having two hosts over which I

Re: [jupyter] About commercial use

Yes, Jupyter is BSD licensed <https://en.wikipedia.org/wiki/BSD_licenses#3-clause_license_.28.22Revised_BSD_License.22.2C_.22New_BSD_License.22.2C_or_.22Modified_BSD_License.22.29> and can be used for commercial purposes. -MinRK On Thu, Jul 28, 2016 at 10:29 AM, 藤岡裕平 wrote: > My name

Re: [jupyter] Can not see the custom.js or log files when using jupyter notebook cmd to launch a IPython notebook

y. This will be overridden by command-line arguments, e.g. jupyter notebook /path/to/notebooks or jupyter notebook --notebook-dir /path/to/notebooks You might verify that this config is picked up, by starting the notebook server with --debug. -MinRK > Thanks in advance! > > 2016-07-28 4:40

Re: [jupyter] using notebook preventing accessing internet?

other than general resource consumption, and idle notebooks shouldn't use much at all. -MinRK On Sun, Jul 17, 2016 at 12:40 AM, Hai Nguyen wrote: > Hi, > > I got very weird issue that I have never met if using Linux. > > I am using notebook on my mac. If keep using the not

Re: [jupyter] JupyterHub fails OAuth redirect with Google Account Chooser

Thanks for following up. I don't really understand how other cookies could confuse the login setup, unless they use the same cookie name as JupyterHub, but at least we now have a lead to follow. -MinRK On Thu, Jul 28, 2016 at 6:17 PM, Tom Lippman wrote: > To follow up in case anyone

Re: [jupyter] Can not see the custom.js or log files when using jupyter notebook cmd to launch a IPython notebook

On Mon, Aug 1, 2016 at 10:16 AM, 薛亚兰 wrote: > Thanks MinRK, found the problem of this notebook.dir issue. > And another problem came when I used the old way as 00-pyspark-setup.py to > start pyspark with Jupyter notebook. We want to combine Spark 1.6.1 and > Jupyter notebook fo

Re: [jupyter] Is it possible to have ipython notebooks of the same Jupyter Notebook server built on one kernel?

a tricky proposition, but might be a great experiment to try as a plugin for the new JupyterLab code, which should support extension in this way much better than the 4.x JavaScript client. -MinRK > > > -- > You received this message because you are subscribed to the Google Groups >

Re: [jupyter] Re: Location of Jupyterhub css for customization?

On Wed, Aug 3, 2016 at 8:02 PM, Ted Liefeld wrote: > Answering my own post (partially) for anyone who comes later > > The hub seems to use this version of the file > > /usr/local/share/jupyter/hub/static/css/style.min.css > That's the default installation location, yes. For customization, on

Re: [jupyter] Re: Location of Jupyterhub css for customization?

On Thu, Aug 4, 2016 at 9:17 PM, Ted Liefeld wrote: > Answering myself again... > > in the hub you can add this to the logout button of the template > > {{user.orm_user.name}} > {{user.name}} works, too. > > > Still working on the docker image to see if it will work there too > > > > > On Thurs

Re: [jupyter] Re: Location of Jupyterhub css for customization?

the system. This is what the SystemUserSpawner does, and one of the reasons I recommend against using the SystemUserSpawner if it can be avoided. Is it important to you for the shell username to be your user's name? -MinRK On Fri, Aug 5, 2016 at 6:48 PM, Ted Liefeld wrote: > Any suggest

Re: [jupyter] How to access the jupyterhub config in a custom authenticator

import Authenticator class MyAuthenticator(Authenticator): thing = Unicode('default', help='helpful info').tag(config=True) then in your jupyterhub_config.py, set this value: c.MyAuthenticator.thing = 'not the default' -MinRK ​ On Fri, Aug 5, 2016 at 11:41 PM, T

Re: [jupyter] How to capture rich display using %%capture or similar magics?

for output-flooding libraries, where you may just want to suppress all the side-effect output on your way to a result, and only look at it if something failed. In this use case, also hiding results wouldn't be desirable, but I can see how the current behavior might be surprising. -MinRK On Tue

Re: [jupyter] How can a cell obtain user name from e.g. HttpContext?

h case the MY_USERNAME env will be available to the server and all of its kernel subprocesses. If the server doesn’t know its name until some time later, you can modify how kernels are started, to pass this environment variable on each kernel launch. -MinRK ​ On Tue, Aug 9, 2016 at 2:19 PM, Paul A wrote:

Re: [jupyter] Interacting with the Jupyter widgets

, rather than triggering events that may or may not do what you want. -MinRK On Fri, Aug 5, 2016 at 11:13 PM, Christopher Brooks wrote: > Hi, > > I'm looking to interact with the jupyter notebook pages in javascript, > largely to do things like send key commands. I've spe

Re: [jupyter] scipy in a jupyter console

stants import g print('after constants import?', hasattr(scipy, 'constants')) And scipy.constants would be defined if *any code* had imported scipy.constants at any time during the program, not just in your current script. So in general, if you want scipy.constants, you should im

Re: [jupyter] How can a cell obtain user name from e.g. HttpContext?

USERNAME env will be available to the server and all >> of its kernel subprocesses. If the server doesn’t know its name until some >> time later, you can modify how kernels are started, to pass this >> environment variable on each kernel launch. >> >> -MinRK >> ​ >

Re: [jupyter] run jupyterhub without system users

d context (remote machines, VMs, docker containers, etc.) in which to run each user's notebook server, via selecting or implementing the appropriate Spawner. -MinRK > > >> We do have some resources on JupyterHub that may be helpful to you: >> >> - JupyterHub

Re: [jupyter] Issues with LDAPauth and home directories

The Authenticator does not affect notebook storage. This is governed by the single-user server started after authentication. The easiest way to do this is if you can mount whatever storage you want to use on the filesystem, and then set that as the notebook_dir for the users. Alternatively, you wou

Re: [jupyter] URL path for extensions on a jupyterhub

Installing and enabling extensions on single-user servers should be all that is needed. That looks like there's an inappropriate absolute URL somewhere. Can you share some more information about your extension? Does it have any explicit URLs, perhaps with `/nbextensions` anywhere? -MinRK O

Re: [jupyter] Who supports https://demohub.jupyter.org? :)

Nobody supports demohub, as it isn't a public service. But I'm usually the one maintaining it. It is an instance of JupyterHub that we occasionally use for presentations, demos, tutorials, testing, etc.. It appears to be working fine right now, though. -MinRK On Thu, Aug 18, 2016

Re: [jupyter] How to let IPython kernel not autostarted in IPython Notebook 3.2.1

I'm not sure disabling autorestart is supported by the notebook application. What prompts your need for dead kernels? -MinRK On Wed, Aug 17, 2016 at 6:16 PM, wrote: > Hi, all > Currently, we are using IPython Notebook + Spark 1.6.1, and I just > found in default situation,

Re: [jupyter] Trouble with setting admin user

You want set(['user']) not set('user'). If you pass a single list to set, you get *each character* as an item in the set, not a single item: In [1]: set('user') Out[1]: {'e', 'r', 's', 'u'} In [2]: set(['user'

[jupyter] traitlets 4.3.0b1

We’ve just pushed a beta for traitlets 4.3. This should be a small release, fixing various bugs and introducing some minor new features. We are doing a beta release because we want to be extra careful that traitlets releases don’t break anything, sinc eit’s t Install it with: pip install --upgra

[jupyter] Re: traitlets 4.3.0b1

Sorry, premature send. We are doing a beta release because we want to be extra careful that traitlets releases don’t break anything, since it's the underpinning of everything. We will release traitlets 4.3 final next week if all goes smoothly with the beta. -MinRK On Fri, Sep 2, 2016 at 1:

Re: [jupyter] Library length limit?

I don't believe there is a size limit, but it's possible there was a race condition only revealed by the larger size and thus longer download time. -MinRK On Tue, Sep 6, 2016 at 4:47 AM, Bruce Sherwood wrote: > Jupyter VPython (pip install vpython; see vpython.org) includes a

Re: [jupyter] Re: Error in embedding jupyter notebooks

If your webapp and the notebook are on different hosts, you will need to relax the notebook server's cross-origin protections: Setting the configuration: NotebookApp.allow_origin='*' will allow any website to launch notebooks. Setting the origin to your other application's domain will allow

[jupyter] traitlets 4.3

We have just released traitlets 4.3, with various fixes and improvements. Highlights include reduced and improved warning and error messages, as well as the ability toaccess the observe/validation dicts via attributes in addition to dict-key access. Change highlights

Re: [jupyter] html rendering

. One workaround you can have is to have relative links to the html files in notebooks with relative URLs, which will resolve to the /files/ version, e.g. [link text](relative/myfile.html) -MinRK ​ On Wed, Sep 7, 2016 at 8:57 PM, Kristen Thyng wrote: > Hi! When I have an html file i

Re: [jupyter] JupyterHub DockerFiler Set Environment Variable not Work

JupyterHub does not pass all of its environment variables to the single-user servers, because it is likely to include hub-level credentials that the user shouldn’t have access to. For this reason, there is a c.Spawner.env_keep configurable that is a whitelist of environment variables that *should*

Re: [jupyter] A Python module with JavaScript and related components

stom location doesn’t mean that that custom location will be found for loading extensions. You can see the locations where Jupyter will look by checking jupyter --paths. The data-paths (+ nbextensions) are where nbextensions will be found. nbextensions_dir should be in one of these unless you are als

Re: [jupyter] JupyterHub: How to persist user data when using GitHub Oauth + DockerContainer?

On Wed, Sep 14, 2016 at 4:23 PM, Kenneth Tran wrote: >1. How can I map `/home/user` in the container to >`some_directory/user` in the host server? > > Assuming you are using DockerSpawner, you would use: c.DockerSpawner.volumes = { '/some_directory/{username}': 'in_container/mountpo

Re: [jupyter] Re: IPython cell magic for C# usage in jupyter notebooks (clrmagic)

python/rmagic.py?at=default&fileviewer=file-view-default> the %%R magic, which handles plotting. -MinRK > > On Wednesday, September 14, 2016 at 11:07:00 PM UTC-5, Lawrence D’Oliveiro > wrote: >> >> On Wednesday, September 14, 2016 at 4:40:04 PM UTC+12, Denis Akhiyarov &

Re: [jupyter] Jupyterhub and multicore system

The most common cause of something like this is an environment variable such as CPU_AFFINITY being set. Such things can pin processes to particular CPU threads. You might check `os.environ` in your notebooks to see if anything suspicious is set that might do something like that. -MinRK On Thu

Re: [jupyter] CSS errors on login page

That does seem very odd. What do you see in `ls -la` in the directory containing this file? -Min On Wed, Sep 21, 2016 at 10:07 PM, Brian Novogradac < brian.novogra...@gmail.com> wrote: > I wanted to make a simple change to the style.min file of the css (color > change). i made the change all wa

Re: [jupyter] IPywidget optional hide close button - Suggestion

It probably makes sense to do something like this in the dashboards view. In the regular notebook view, I'm not sure we would want users to not have access to closing things, though. https://github.com/jupyter-incubator/dashboards On Wed, Sep 21, 2016 at 12:24 AM, oscar6echo wrote: > Hi, > > I

Re: [jupyter] jupyterhub and quota

You will only get a 500 in the JupyterHub page if it fails to start the notebook server at all, so there's no server running to prompt the user to take any actions. Can you tell which page is getting a 500 error? -Min On Fri, Sep 30, 2016 at 8:07 AM, Marcin Kostur wrote: > Hi, > > When user tri

Re: [jupyter] SSL certificate errors from Jupyter/Docker

That's exceedingly strange, and I have not seen it before. The proxy is involved when your browser connects to the notebook server, but it shouldn't be involved at all when the kernel connects to the outside world. Can you check `os.environ` to see if there are any relevant environment variables th

Re: [jupyter] Re: Numpy in Jupyter GCC-4.6 error

Can you compare os.environ in the two contexts? And `sys.executable` and `sys.path`? -Min On Sun, Oct 2, 2016 at 10:20 PM, Peter Sprague < geovisionenvironmen...@gmail.com> wrote: > Setting LD_LIBRARY_PATH had worked with Python 2.7 in the past. Doesn't > help now. > > https://groups.google.com

Re: [jupyter] Propose change in how frontends handle payloads

others). Summarizing some discussion from the dev meeting: I support deprecating the pager in JupyterLab, and moving IPython’s default behavior of ? to produce regular output (I’ve used extensions <https://github.com/minrk/ipython_extensions/blob/master/extensions/print_page.py> to do this

Re: [jupyter] SSL certificate errors from Jupyter/Docker

What do you see in !curl -v https://github.com > /dev/null ? It’s certainly odd that requests is connecting to something funny. ​ On Mon, Oct 3, 2016 at 10:37 PM, Ted Liefeld wrote: > Min > > here is what I see > > environ({ > 'HOSTNAME': '0955ba141195', > 'GS_CODE_ROOT_DIR': '/combined', > '

Re: [jupyter] Jupyterhub and multicore system

ronment variable >> such as CPU_AFFINITY being set. Such things can pin processes to particular >> CPU threads. You might check `os.environ` in your notebooks to see if >> anything suspicious is set that might do something like that. >> >> -MinRK >> >> On T

Re: [jupyter] Processing mouse events

It isn't generally possible to know that the end of the buffer has been reached, because these events can just keep coming. There are some options, though: 1. you could set a small delay on handling the event, and restart the counter on each new event. That way, your response is limited to ~50ms,

Re: [jupyter] Executing script statement-by-statement

I don't believe any protocol changes are needed for this. You can already run a script through Jupyter, breaking it up however you want. The trick would be writing a runner for a given script (you would need to write a different one for every language to do the parsing correctly), and then submit t

Re: [jupyter] Can we use Jupyter on Linux Power platform

I imagine it should work well enough. If you have Python, most Jupyter dependencies should be relatively easy to install. I'd start with `pip install jupyter`, and see how far you get. It's possible some of the compiled dependencies need a little help getting built, but those are usually issues you

Re: [jupyter] JupyterLab Migration

This means that really all Jupyter javascript is actually called "juptyerlab" now, yes? Any new Jupyter js would go on the jupyterlab org, even if it's not to do with jupyterlab (e.g. services)? -Min On Thu, Oct 13, 2016 at 3:20 PM, Steven Silvester wrote: > Hi all, > > Yesterday we moved Jupyt

Re: [jupyter] JupyterLab Migration

er aspects we missed? > Makes sense! I like the namespace effects of orgs. -Min > > Cheers, > > Brian > > On Fri, Oct 14, 2016 at 8:41 AM, MinRK wrote: > > This means that really all Jupyter javascript is actually called > > "juptyerlab" now, yes? Any

Re: [jupyter] JupyterLab Migration

. There's a threshold where following all the repos on an org stops being feasible, and orgs can reduce the effective number of 'things' to keep track of by putting them in buckets. > > On Fri, Oct 14, 2016 at 9:51 AM, MinRK wrote: > > > > > > On Fri, Oct 14, 201

Re: [jupyter] How to make CFFI stdout print in jupyter?

You can use a tool called wurlitzer <https://github.com/minrk/wurlitzer#wurlitzer> to capture C-level piped output: from wurlitzer import sys_pipeswith sys_pipes(): call_c_function() -Min ​ On Mon, Oct 17, 2016 at 10:46 AM, Dmitry Ulyanov < dmitry.ulyanov@gmail.com> wrote:

Re: [jupyter] Notebook/Lab team structure (while we finalize the rest)

On Fri, Oct 21, 2016 at 8:57 AM, Fernando Perez wrote: > Hi all, > > we've had discussions since the last dev meeting about organizing the > project into more well-defined teams for each major area, since we're well > past the point where "everyone does everything" can remotely work. > > While we

Re: [jupyter] Keyboard input

Yes! You need to call keyboard_manager.register_events(element);` where element is the element you want to get its own keyboard events, and keyboard_manager is accessible as either Jupyter.notebook.keyboard_manager or this.keyboard_manager, depending on context. -Min ​ On Tue, Oct 18, 2016 at 9

Re: [jupyter] Jupyterhub and SSL

On Sat, Oct 29, 2016 at 9:36 AM, 'csw81' via Project Jupyter < jupyter@googlegroups.com> wrote: > I have been using Jupyter Notebook for a long time already. Jupyter > Notebook is really great. Thank you all! > > Now, I would like to run a jupyterhub server (using Ubuntu 16.04). This > works as lo

Re: [jupyter] LDAP ja home directory creation

You might want to override Spawner.start to do the actions you are looking for when a user’s server starts. from jupyterhub.spawner import LocalProcessSpawnerclass MySpawner(LocalProcessSpawner): def start(self): setup_home_directory(self.user.name) # < define this function ret

Re: [jupyter] download is .ipynb in safari

On Sat, Oct 29, 2016 at 3:39 AM, Matthias Bussonnier < bussonniermatth...@gmail.com> wrote: > Thanks for the report, > > Unfortunately as far as I remember there is nothing we can do about this. > Safari is known to have many problems with Jupyter Notebook in general, > and is not one of the brows

Re: [jupyter] download is .ipynb in safari

This PR <https://github.com/jupyter/notebook/pull/1861> fixes the file-extension-adding on Safari by declaring the correct mime-type. -MinRK ​ On Mon, Oct 31, 2016 at 3:52 PM, MinRK wrote: > > > On Mon, Oct 31, 2016 at 3:45 PM, Thomas Kluyver wrote: > >> On 31 Octo

Re: [jupyter] download is .ipynb in safari

On Mon, Oct 31, 2016 at 3:45 PM, Thomas Kluyver wrote: > On 31 October 2016 at 13:24, MinRK wrote: > >> it should probably be something notebook-specific, such as >> `application/vnd.jupyter.notebook+json`. >> > > We already decided on application/x-ipynb+json as t

Re: [jupyter] How to configure JupyterHub + DockerSpawner + PAM Authentication with username containing the @ sign

What do you see in the JupyterHub logs and the logs of the docker container that failed to start? -Min On Tue, Nov 1, 2016 at 9:09 AM, Amichai Rotman wrote: > Hi Group, > > I am racking my brain for a week now: > > I am trying to install JupyterHub on a CemtOS 7 server. I want it to > authentic

Re: [jupyter] Figuring out Jupyter Notebook config

nd dumps its config instead of starting. We should probably add a similar config option to the base Application so that every descendant can easily do this. -MinRK ​ On Mon, Oct 31, 2016 at 11:32 PM, Kyle Kelley wrote: > Does anyone know of a good way to dump the config of a notebook server

Re: [jupyter] Re: Jupyterhub + docker + swarm + can't connect to kernel - How to debug?

Ted, Sorry we didn't get to this before you did, a lot of Jupyter folks are at a Jupyter Team Meeting all week, so are falling behind on GitHub and email. Thanks for posting your solution! -MinRK On Tue, Nov 8, 2016 at 11:25 AM, Ted Liefeld wrote: > Found a solution and posting it

Re: [jupyter] Problem: pip install jupyter hangs at tornado installation.

You can try getting more output from pip as it works with: pip install -vv -MinRK On Sat, Nov 12, 2016 at 12:43 PM, Yuval Langer wrote: > Trying to install jupyter using pip, I find it hangs at tornado's > installation: > > $ pip3 install --user --upgrade jupyter >

Re: [jupyter] Browser communication with Jupyter Python kernel

JupyterLab services is for use in the browser, that’s what JupyterLab uses to talk to these APIs. If you are already in the existing pre-Lab notebook and want to run code on the current kernel, the API you are after is: Jupyter.notebook.kernel.execute("some code", callbacks) -MinRK

Re: [jupyter] Help needed: creating a notebook extension

You can publish Javascript from the kernel with: from IPython.display import display, Javascript display(Javascript("""your_js()""")) That might be what you need. -MinRK ​ On Fri, Nov 4, 2016 at 3:22 PM, Logan Page wrote: > I have more or less figured out w

[jupyter] JupyterHub 0.7.0b1

We have just released the first beta of JupyterHub 0.7. Please take it for a spin: pip install --upgrade --pre jupyterhub Main new features: - A Services API for JupyterHub to manage adding additional processes with API access to the Hub, including running additional web services behind

Re: [jupyter] Re: JupyterHub 0.7.0b1

On Sun, Nov 13, 2016 at 7:31 PM, Jeremy Douglass wrote: > Regarding: > > The single-user server is now provided by the jupyterhub package > > > Now that JupyterHub provides the single-user server, (how) will the hub > work with Jupyter Docker Stacks (https://github.com/jupyter/docker-stacks) > go

Re: [jupyter] Does JupyterHub work with jupyter-incubator's dashboards nbextension?

Any nbextension should work with JupyterHub. You may need to do a system-wide install/enable of the extension for it to work for all users. -MinRK On Wed, Nov 16, 2016 at 1:08 PM, Pav A wrote: > https://github.com/jupyter-incubator/dashboards > > E.g. see http://jupyter.cl

Re: [jupyter] Re: JupyterHub 0.7.0b1

On Tue, Nov 15, 2016 at 11:21 PM, Jeremy Douglass wrote: Thank you, Min RK! > > Would doing that (e.g. me modifying the DockerFile and rebuilding the > image for Jupyter DataScience) be necessary if I today I wanted to run a > multi-user JupyterHub in which the hub hosted individual Datascience >

Re: [jupyter] jupyterhub+docker issue with change of user password

Can you run `docker exec` to check that the contents of the passwd file are still in-sync or not after the update? There's a chance that mounting the file doesn't quite work (depending on how the file is modified), and you may need to mount a directory, instead to ensure that files are correct. -

Re: [jupyter] JupyterHub : Inserting data into the hub DB using API and HTML form

On Wed, Nov 16, 2016 at 9:32 PM, Joshua Dunham wrote: Hi Jupyters, > > I'm working through setting up JupyterHub and am trying to think > through how to provide a few group of people access to a single hub. Some > questions: > > - I've noticed that {username} is a variable I can use in some

Re: [jupyter] Browser communication with Jupyter Python kernel

On Sat, Nov 12, 2016 at 11:39 PM, Christopher Brooks wrote: Wonderful, thanks very much MinRK. Is callbacks just a function which takes > a single parameter, or are there docs on this api which I have overlooked? > callbacks is a structure for different types of callbacks. Its shape is des

Re: [jupyter] Problem: pip install jupyter hangs at tornado installation.

That looks like something is wrong with either pip, PyPI, or the network, because it's stuck trying to download the package index, a simple, small text file. Maybe clear the pip cache? Can you install other packages? -Min On Sun, Nov 13, 2016 at 8:00 AM, Yuval Langer wrote: > Yep. The process i

Re: [jupyter] Browser communication with Jupyter Python kernel

On Fri, Nov 18, 2016 at 2:05 PM, Doug Blank wrote: On Friday, November 18, 2016 at 7:52:23 AM UTC-5, Min RK wrote: >> >> On Sat, Nov 12, 2016 at 11:39 PM, Christopher Brooks >> wrote: >> >> Wonderful, thanks very much MinRK. Is callbacks just a function which &g

Re: [jupyter] JupyterHub : Inserting data into the hub DB using API and HTML form

eation, Strategic, Learner > > Sent from Nylas N1 > <https://link.nylas.com/link/68p140bimyedpd78wkdi7pwew/local-027c75cc-cc2f/0?redirect=https%3A%2F%2Fnylas.com%2Fn1%3Fref%3Dn1&r=anVweXRlckBnb29nbGVncm91cHMuY29t>, > the extensible, open source mail client. > > On Nov 18 2016, a

Re: [jupyter] jupyterhub centos 7

On Mon, Nov 21, 2016 at 6:31 AM, Lak Sa wrote: *hi guys,* > > *i found an issue - I am trying to install jupyterhub on centos 7.* > > *this instruction: https://github.com/jupyterhub/jupyterhub > * > > *ask for apt-get. * > *i realise i need to install RP

Re: [jupyter] How to link and render a markdown file readme.md from a notebook MyNotebook.ipynb?

There isn't a handler that renders markdown (the server does no markdown rendering), so there is no URL for such a thing unless you add an extension that provides it. JupyterLab adds rendering support for markdown files in the editor, though. -Min On Tue, Nov 15, 2016 at 7:55 PM, Owen Kelly wrot

[jupyter] PyCon tutorial

Carol brought up the idea that an IPython and/or Jupyter tutorial at PyCon would be a good idea, since we didn't do one last year. The deadline for tutorials in *next week, 30 November* We have a few tutorials available, which we can derive from what we've done in the past: - classic "IPython in

Re: [jupyter] JupyterHub KubeSpawner non root

I think you'll want to use Authenticator.pre_spawn_start to set spawner.uid (and pot

Re: [jupyter] Help needed: SSL certificates showing up in Jupyter Notebook

Hi! I’m not sure what you mean by “the certificates are showing up in the notebook.” Do you mean the files are visible in the file-listing (dashboard) view? If that’s the case, you can put the certificate and key files outside of the notebook directory. By default, the notebook server starts in th

Re: [jupyter] PyCon tutorial

onnier < bussonniermatth...@gmail.com> wrote: > Hey Min, > > > On Wed, Nov 23, 2016 at 12:11 PM, MinRK wrote: > > Carol brought up the idea that an IPython and/or Jupyter tutorial at > PyCon > > would be a good idea, since we didn't do one last year. The

Re: [jupyter] Re: Redirect problem with Jupyterhub

On Tue, Nov 29, 2016 at 10:19 PM, Asher Wasserman wrote: > It seems the new 0.7.0b1 release is now working for me with Firefox. I > can't find the exact issue that was causing it, but the change log mentions > some redirect-loop fixes. > Several of the bugfixes relate to issues that could cause

[jupyter] nbformat 4.2

useful in git hook workflows, etc. See the changelog <http://nbformat.readthedocs.io/en/latest/changelog.html#id1> or GitHub <https://github.com/jupyter/nbformat/issues?q=milestone%3A4.2+is%3Aclosed> for more details. Thanks everyone! -MinRK ​ -- You received this message bec

[jupyter] [ANN] Release jupyterhub 0.7

We’ve just released JupyterHub 0.7! The highlight is a new Services API for running additional services behind JupyterHub, which can use the Hub for authentication. Plus, loads of other fixes and improvements, especially resolving various

[jupyter] Jupyter HPC mailing list

Based on conversations today with computing institutes here in France following JupyterDay Paris, and the growing community of people deploying and using Jupyter in HPC/cluster contexts, we've created a mailing list dedicated to using Jupyter in HPC: https://groups.google.com/forum/#!forum/jup

[jupyter] Re: Jupyter HPC mailing list

Oops, I made it a private group. It should be public, now. -Min On Wed, Dec 7, 2016 at 12:12 PM, MinRK wrote: > Based on conversations today with computing institutes here in France > following JupyterDay Paris, and the growing community of people deploying > and using Jupyter in HP

Re: [jupyter] Opting out of readthedocs advertisements

s possible. I'm not quite sure how exactly that would work, since RTD doesn't seem to have org accounts. Since we use a custom theme on the main jupyter.readthedocs.io at the moment, that site is not currently supporting RTD since ads are not shown. -MinRK On Fri, Dec 9, 2016 at 9:2

[jupyter] [ANN] nbconvert 5.0

We’ve just released nbconvert 5.0, with lots of nice fixes and improvements. Among the key changes are switching the default latex engine from pdflatex to xelatex, which should get much better unicode and font handling, among other things. There is also a new ASCIDoc Exporter, and loads of other bu

[jupyter] [ANN] nbdime 0.1

We’ve made our first release of nbdime : tools for diffing and merging notebooks. pip install --upgrade nbdime Key features: - nbdiff for diffing notebooks in the terminal - nbdiff-web for viewing a rich, rendered diff of two notebooks - nbmerge

Re: [jupyter] connection to the notebook server could not be established

Chrome results in the same thing? It seems like the cookie setting has to be failing somehow, but I see the cookie in the headers. Can you reproduce this in a Private Browsing or Incognito window? -Min On Mon, Dec 19, 2016 at 11:38 AM, Thomas Kluyver wrote: > Can you check in the browser to see

Re: [jupyter] 'real-time' bqplot update from external source?

If you fetch data in a background thread and update the bqplot object, this should work. For the background updates to work fully, the most recent version of ipykernel may be needed, so make sure everything is up-to-date. -Min On Mon, Dec 19, 2016 at 11:31 AM, wrote: > Hi all, > Is there a way

Re: [jupyter] Auto restart long-running kernels

On Fri, Dec 9, 2016 at 3:09 PM, Jesus Castello wrote: Hello, > I want to run a Jupyter notebook server for my students. My only concern > is if someone runs code that takes too long to finish (or even worse, an > infinite loop). > > Is there a way to make a kernel auto restart after 10 seconds of

Re: [jupyter] How to configure/Integrate PySpark with Jupyter on Windows

The default IPython kernel doesn't set up a spark context by default. There are lots of ways to hook Python up to Spark, so I'm not sure which you have tried, but it is quite possible that the Spark integration is done through a custom Kernel spec, which you may find in the Kernel drop-down menu.

Re: [jupyter] Internal Server Error with SystemUserSpawner

Can you show the output of docker logs jupyter-dbiringer? That might be where the 500 error came from. -Min ​ On Tue, Dec 20, 2016 at 1:55 AM, Dennis Biringer wrote: > Hi, > > I've got a configuration of JupyterHub that uses the SystemUserSpawner. > After I log in, I hget '500 : Internal Server

Re: [jupyter] Using batchspawner while not running as root

On Wed, Dec 21, 2016 at 2:47 PM, Andreas Hilboll wrote: > Dear Jupyter community, > > I'm planning to set up a jupyterhub instance on our HPC cluster, where > individual notebooks should be spawned on the compute nodes via SLURM. > > From what I'm reading, the batchspawner would be my way to go.

Re: [jupyter] [REL] matplotlib 2.0.0rc2

This is awesome. Very excited for MPL 2.0. On Wed, Dec 21, 2016 at 6:06 AM, Thomas Caswell wrote: > Folks, > > We are happy to announce matplotlib v2.0.0rc2 ! > > Please re-distribute this widely. > > This is the final planned release candidate for the long awaited mpl v2.0 > release. > > For th

Re: [jupyter] GitHub Automation, migrating issues.

Very cool, thanks Matthias! On Tue, Dec 27, 2016 at 2:50 PM, Matthias Bussonnier < bussonniermatth...@gmail.com> wrote: > Hi all, > > [Cross posting IPython-dev and Jupyter Google group] > > Hope you are enjoying the lasts few days of 2016 ! > > I've been recently working on GitHub automation, in

Re: [jupyter] Jupyter 0.7.0 - errors adding single-letter users when I start up a clean install of jupyterhub

On Tue, Jan 3, 2017 at 6:21 AM, Jonathan Morgan < jonathan.morgan@gmail.com> wrote: Hello, > > I am having an odd problem with jupyterhub on two ubuntu machines where I > just installed jupyterhub - one is Ubuntu 16.04, one is ubuntu 16.10. > Jupyterhub version is 0.7.0. Python version is the

  1   2   3   >