Ipython

Now Jupyter

Try online Jupyter

or https://try.jupyter.org/

(based on tmpnb

)

Jupyter talk by Fernando Perez (Video

+Slides

Install python2 and python3 notebooks

[+]

Installation based in pip2 and pip3. First we must sure to use the unpackaged versions of pip

# apt-get remove python-pip python2-pip

.
Then

# aptitude install python-setuptools python3-setuptools python-dev python3-dev
# easy_install-2.7 pip
# easy_install3 pip

We must now install jupyter < 6 with support for python2 kernel

# pip3 install notebook

To install the python2 kernel

# pip2 install ipykernel

And each user must activate the kernel at user level:

$ python2 -m ipykernel install [--user]

Without the --user option the installation is sytem wide
Once

$ jupyter notebook

is launched, the two kernel could appear as available.

Install jupyterhub

[+]

Install and use according to the following script (https://github.com/jupyterhub/jupyterhub/blob/master/README.md)

jupyterhub.sh

#!/usr/bin/env bash
#=== INSTALLATION ===
# # pip3 install --upgrade notebook
# JupyterHub can be installed with pip, and the proxy with npm:
# # npm install -g configurable-http-proxy
# # pip3 install jupyterhub

#=== CERTIFICATE GENERATION: ===
# From: https://workaround.org/ispmail/jessie/create-certificate
# # openssl req -newkey rsa:4096 -nodes -sha512 -x509 -days 3650 -nodes -out /etc/ssl/certs/mailserver.pem -keyout /etc/ssl/private/mailserver.pem
# On "Common Name" use the hostname, e.g, www.yourserver.com

#=== LAUNCH COMMAND as root: === 
jupyterhub --port 4443 --ssl-key /etc/ssl/private/mailserver.pem --ssl-cert /etc/ssl/certs/mailserver.pem

Install jupyter extensions:

[+]

Follow the instructions here

and later here

Install (old)

[+]

The official updated documentation: Jupyter Kernels

.

To global install work in

/usr/local/share/jupyter/kernels

or something like this. The examples below are local installations. If the kernel is official can be installed as

python2 -m IPython kernelspec install-self

python3 -m IPython kernelspec install-self

See discussion here

. Or this

other thread

Cita:

You can perfectly run IPython 3 with Python 3 and kernels on Python 2.
for the time being you will need to install IPython notebook with python 2 and issue a

$ ipython2 kernelspec install-self

This will make the IPython2 kernel available with IPython3 server running on python 3.

See kernels

$ cat ~/.ipython/kernels/ir/kernel.json
{"argv": ["R","-e","IRkernel::main()","--args","{connection_file}"],
 "display_name":"R"
}

$ cat ~/.ipython/kernels/python2/kernel.json
{
    "display_name": "Python 2",
    "language": "python",
    "codemirror_mode": {
	"version": 3,
	"name": "ipython"
    },
    "argv": [
	"/usr/bin/python",
	"-c",
	"from IPython.kernel.zmq.kernelapp import main; main()",
	"-f",
	"{connection_file}"
    ]
}

o mejor el oficial de http://try.jupyter.org

jovyan@374b5bfac0c6:~$ cat /usr/local/share/jupyter/kernels/python2/kernel.json                                                           
{                                                                                                                                         
 "display_name": "Python 2",                                                                                                              
 "language": "python",                                                                                                                    
 "argv": [                                                                                                                                
  "/usr/bin/python2",                                                                                                                     
  "-m",                                                                                                                                   
  "IPython.kernel",                                                                                                                       
  "-f",                                                                                                                                   
  "{connection_file}"                                                                                                                     
 ]                                                                                                                                        
}

Parallel Ipython

[+]

https://fisica.udea.edu.co:8080/notebooks/prog/2014/T12A/ParaleloSimple.ipynb

Ipython notebook with pylab support

[+]

$ ipython notebook --InlineBackend.figure_format=svg

and in the first Ipython cell:

%pylab inline

Ipython notebook server

[+]

Fast way

[+]

Listen all the interfaces

ipython notebopok --ip=*

How to setup an ipython notebook server with the '%pylab inline' already active in Debian Wheezy. We assume that you have a full SciPy? installation as expalined in ubuntu.sh.

Instructions based in this tutorial

Configuring the notebook server

First, get your password hashed with IPython utility (From the terminal console!).

In [1]: from IPython.lib import passwd
In [2]: passwd()
Enter password:
Verify password:

Next, create a new IPython configuration profile.

$ ipython profile create myserver
$ vim ~/.ipython/profile_myserver/ipython_notebook_config.py

Uncomment and edit the following lines in the IPython profile config file - ~/.ipython/profile_myserver/ipython_notebook_config.py

c = get_config()

c.IPKernelApp.pylab = 'inline'
c.NotebookApp.ip = '*'
c.NotebookApp.open_browser = False
c.NotebookApp.password = u'sha1:yourhashedpassword'
c.NotebookApp.port = 9999

In order to guarantee a secure conection a certificate must be created

$ openssl req -x509 -nodes -days 365 -newkey rsa:1024 -keyout mycert.pem -out mycert.pem

Running the server: now, you can run the server, to keep it running, you can manage the process with Supervisor.
```
$ ipython notebook --profile=myserver --certfile=mycert.pem
```
To check if inline chart works, just type that line in your new notebook at https://your.site:9999 (note the https protocol)
```
In [1]: plot(arange(10))
Out[1]:
[<matplotlib.lines.Line2D at 0x39c8e50>]
```

That’s it, now you are able to access your notebook from everywhere.

The path need to be relative to where the server has been started, and prefixed with files/. e.g:

[PDF](files/workingdir/output.pdf)

Ipython notebook: make output cells like markdown

[+]

from IPython.core.display import HTML
def foo():
    raw_html = "<h1>Yah, rendered HTML</h1>"
    return HTML(raw_html)

foo()

See: http://stackoverflow.com/a/13770691

Ipython notebook: display HTML code (as a cell output):

[+]

See http://stackoverflow.com/a/13770691/2268280

from IPython.core.display import HTML
def foo():
    raw_html = "<h1>Yah, rendered HTML</h1>"
    return HTML(raw_html)

foo()

or simply:

from IPython.core.display import HTML
raw_html = "<h1>Yah, rendered HTML</h1>"
HTML(raw_html)

Yah, rendered HTML

Save notebook as LaTeX/PDF

[+]

File -> Download as reST (.rst)

in your terminal

$ pandoc -s file.rst -t latex -o file.tex

If necesary:

$ sudo apt-get install pandoc

Install in Debian 7 (Wheezy)

[+]

In Debian 8 use pip3 and the default instructions.

Install npm from here

# curl -sL https://deb.nodesource.com/setup | bash -
# apt-get install -y nodejs
# #or directly from: http://gfif.udea.edu.co/downloads/nodejs_0.10.36-1nodesource1~wheezy1_amd64.deb

and follow the instructions

# sudo apt-get install npm nodejs-legacy
# npm install -g configurable-http-proxy
# pip install ipython --upgrade
# pip install ipython --upgrade
# pip install pyzmq --upgrade
# pip install tornado --upgrade  
# pip install jsonschema

Use Ipython Cell as optional editor

[+]

The Ipython Cells have its own sys.argv. The following check for Ipython before call sys.argv

#######IPYTHON CELL #######
[#]%%writefile run.py
#!/usr/bin/env python
CODE
def run_from_ipython():
    try:
        __IPYTHON__
        return True
    except NameError:
        return False

if not run_from_ipython() and len(sys.argv)>1:
    CODE

El documento original está disponible en https://clustercien.udea.edu.co/web/tiki-index.php?page=Ipython