Last week, I wrote a guest blog post at Bad Hessian about how to use IPython Notebook along with Amazon EC2 as your data science & analytics platform. I won’t reproduce the whole article here, but if you are interested in step-by-step instruction on how to setup an Amazon EC2 instance to use IPython Notebook, see the SlideShare presentation below which outlines the steps needed to setup a remote IPython Notebook environment (or, PDF download).
If you already have experience setting up EC2 images and just need the IPython Notebook settings, here are the commands that are needed to set up your IPython public notebook server.
#### Start IPython, generate SHA1 password to use for IPython Notebook server
$ipythonPython2.7.5|Anaconda1.8.0(x86_64)|(default,Oct242013,07:02:20)Type"copyright","credits"or"license"formoreinformation.IPython1.1.0--AnenhancedInteractivePython.?->IntroductionandoverviewofIPython's features.%quickref -> Quick reference.help -> Python'sownhelpsystem.object?->Detailsabout'object',use'object??'forextradetails.In:fromIPython.libimportpasswdIn:passwd()Enterpassword:Verifypassword:Out:'sha1:207eb1f4671f:92af695...'#### Create nbserver profile
$ipythonprofilecreatenbserver[ProfileCreate]Generatingdefaultconfigfile:u'/.ipython/profile_nbserver/ipython_config.py'[ProfileCreate]Generatingdefaultconfigfile:u'/.ipython/profile_nbserver/ipython_qtconsole_config.py'[ProfileCreate]Generatingdefaultconfigfile:u'/.ipython/profile_nbserver/ipython_notebook_config.py'[ProfileCreate]Generatingdefaultconfigfile:u'/.ipython/profile_nbserver/ipython_nbconvert_config.py'#### Create self-signed SSL certificate
$opensslreq-x509-nodes-days365-newkeyrsa:1024-keyoutmycert.pem-outmycert.pem#### Modify ipython_notebook_config.py configuration file
#### Add these lines to the top of the file; no other changes necessary
#### Obviously, you'll want to add your path to the .pem key and your password
# Configuration file for ipython-notebook.
c=get_config()# Kernel config
c.IPKernelApp.pylab='inline'# if you want plotting support always
# Notebook config
c.NotebookApp.certfile=u'/home/ubuntu/certificates/mycert.pem'c.NotebookApp.ip='*'c.NotebookApp.open_browser=Falsec.NotebookApp.password=u'sha1:207eb1f4671f:92af695...'# It is a good idea to put it on a known, fixed port
c.NotebookApp.port=8888#### Start IPython Notebook on the remote server
Lately, I’ve been using Jupyter Notebooks for all of my Python and Julia coding. The ability to develop and submit small snippets of code and create plots inline is just so useful that it has broken the stranglehold of using an IDE while I’m coding. However, the one thing that was missing for a smooth transition was line numbers in the cells; luckily, this can be achieved in two ways.
The easiest way to add line numbers to a Jupyter Notebook is to use the keyboard shortcut, which is Ctrl-m to enter Command Mode, then type L. Just highlight the cell you are interested in adding line numbers to, then hit the keyboard shortcut to toggle the line numbers.
Add Line Numbers to All Cells at Startup
While the keyboard shortcut is great for toggling line numbers on/off, I prefer having line numbers always on. Luckily, the IPython Dev folks on Twitter were kind enough to explain how to do this:
@randyzwitch add `IPython.Cell.options_default.cm_config.lineNumbers = true;` to your custom.js
Edit 11/4/2015: Thanks to reader Nat Dunn, I’ve been made aware that the above method no longer works, which isn’t a surprise given the amount of changes between IPython Notebook to the entire Jupyter project in the past 2 years.
Version 1.2 of the RSiteCatalyst package to access the Adobe Analytics API is now available on CRAN! Changes include:
Removed RCurl package dependency
Changed argument order for GetAdminConsoleLog to avoid error when date not passed
Return proper numeric type for metric columns
Fixed bug in GetEVars function
Added validate:true flag to API to improve error reporting
Removed remaining references to Omniture
For the most part, the only noticeable change for most users will be that you no longer need to call as.numeric() on a DataFrame after getting the results of an API call, as all functions now return the proper numeric type.
Changes from Development Version
For any of you out there that may have installed the 1.2 development version directly from GitHub, the only difference between the 1.2 development version and the stable, CRAN version of the package is that support for the Adobe Analytics Real Time API has been removed. This functionality will continue to be developed on the 1.3 development branch on GitHub.
For this release, I’ve made a more concerted effort to test RSiteCatalyst on various platforms outside of OSX (where I do my development). RSiteCatalyst works in the following environments:
OSX Lion and prior
Ubuntu 12.04 LTS
Windows 7 64-bit SP1
Windows 8.1 64-bit
R 2.15.2 and newer
R and RStudio
If your environment is not listed above, it is still likely the case that RSiteCatalyst will work in your environment, as there is no operating-system-specific code in the package. If you are finding issues, validate that you have all package dependencies installed, your Adobe account has Web Service Access privileges (set in Admin panel), you have permission access to the report suites you are trying to access (also an Admin panel setting) and that your company doesn’t have any firewall settings that would prevent API access.
If you run into any problems with RSiteCatalyst, please file an issue on GitHub so it can be tracked properly. Note that I’m not an Adobe employee, so I can only provide so much support, as in most cases I can’t validate your settings to ensure you are set up correctly (nor do I have any inside information about how the system works 🙂 )