Installing IPython Notebook with Jupyter on OSX: A Step-by-Step Guide for Data Scientists

====================================================================================

In this article, we will walk through the process of installing IPython Notebook (now known as Jupyter) along with its corresponding R kernel. This comprehensive guide is specifically tailored for data scientists and analysts who work primarily on macOS systems.

Prerequisites & Preliminary Steps

------------------------------

Before diving into the installation procedure, ensure that you have met all the following requirements:

1. You are running OSX 10.9 (Mavericks) or later.

2. Python has been installed on your system (preferably version 3.x).

3. pip and/or conda package managers should be available.

Step-by-Step Installation Guide

------------------------------

### Step 1: Install IPython/Jupyter Notebook

To begin, open a terminal window from the Applications/Utilities folder or use Spotlight to search for 'Terminal.' Once you have successfully launched Terminal, navigate to your desired working directory using `cd` followed by the path.

Next, run one of the following commands depending on whether you prefer pip (the Python package manager) or conupyter (a command-line tool provided with Jupyter):

Using pip:

```bash

pip install jupyter notebook ipython_kernel_launcher

```

Alternatively,

Using conda (optional but recommended for managing scientific packages):

1. Install Anaconda/Miniconda from the official 'Anaconda website'(https://www.continuum.io/downloads#anaconda).

2. Once installation is complete, navigate to your Minicona‌nda or Anaconda3 directory in Terminal.

3. Activate the respective virtual environment (e.g., `source activate anaconda3` on Mac OS X) if needed.

Then execute:

```bash

conda install -c https://github.com/jupyter/ IPython jupyter_kernel_gateway ipykernel ipywidgets pandas scikit-learn numpy scipy matplotlib seaborn jupyterlab

```

### Step 2: Install R and devtools

Launch Terminal again, then navigate to the directory where you wish to store your Jupyter notebooks. Next, run the following command in R:

```R

# Install rmarkdown package from CRAN if not already installed.

install.packages('rmarkdown')

# Now install IRkernel for IPython/Jupyter Notebook using devtools.

devtools::install_github('IRIS-SHEfLoaS/ IRKernel')

```

Verify that the installation was successful by restarting your R environment and checking for any errors or warnings:

```R

library(IRkernel)

ir() # This should display information about IPython kernel.

```

### Step 3: Launch Jupyter Notebook & Test R Kernel

To launch the Jupyter notebook, simply execute the following command in Terminal from within your working directory where you installed all required packages earlier:

```bash

jupyter nbextension install ipython_contrib.nbextensions # Optional - enables additional features like line magics.

jupyter kernel list # Verify that 'IR' (R) appears among available kernels.

# To start a new Jupyter notebook with R kernel, type:

jupyter notebook --kernel ir --ip=127.0.0.1:8085

```

You should now see an IPython Notebook interface open in your default web browser at 'http://localhost:8085'(http://localhost:8085) or any other specified port number (in this case, 8085). Click on 'New' > 'Python3 (IR)' to create a new notebook using the R kernel.

Troubleshooting Tips

----------------

If you encounter issues during installation or execution of Jupyter Notebook with its associated R kernel, refer to these common solutions:

1. Make sure that both Python and pip/conda are updated.

2. Check for any conflicts between different versions of IPython installed via pip and conda by running `pip list` and/or `conda info -f ipykernel`.

3. Try reinstalling the required packages using `pip uninstall --force 'package_name' && pip install 'package_name'`.

-------

With these step-by-step instructions, you should now have a fully functional IPython Notebook environment with an R kernel installed on your OSX system. This powerful combination will enable data scientists and analysts to easily transition between Python code execution, interactive visualizations using popular libraries like pandas, NumPy, scikit-learn, matplotlib, etc., while still benefiting from the expressive power of R programming language.

Happy coding!

R install ipython notebook