This is a collection of notebooks and other resources published in the Unidata blogs. In addition to viewing them on the blog, you can also load them up in nbviewer.

The current list of dependencies for this collection of notebooks is:

• numpy
• netcdf4-python
• matplotlib
• cartopy
• siphon
• MetPy

The easiest way to install these libraries is with conda.

1. Install Miniconda (Python 3.4) from Continuum Analytics. (Determine if your OS 32 or 64 bit)
2. Once Miniconda is installed, from the command line (e.g., OS X terminal, cmd.exe), run these instructions to clone the repository and create the environment:

git clone https://github.com/Unidata/blog-notebooks

cd blog-notebooks

conda env create -f environment.yml

If your default shell is NOT bash, first type bash. To activate or switch to a conda environment, you can source activate <environment>. For example,

source activate unidata-blog

To switch and/or deactivate environments:

source deactivate
source activate <environment>

To activate or switch to a conda environment, you can activate <environment>. For example,

activate unidata-blog

To switch and/or deactivate environments:

deactivate
activate <environment>

From the blog-notebooks directory and with the unidata-blogs environment active, run:

jupyter notebook

This starts the webserver for running jupyter notebooks. You should see some output like the following:

[I 11:59:31.409 NotebookApp] Serving notebooks from local directory: /Users/rmay/repos/blog-notebooks
[I 11:59:31.409 NotebookApp] 0 active kernels
[I 11:59:31.409 NotebookApp] The Jupyter Notebook is running at: http://localhost:8888/
[I 11:59:31.409 NotebookApp] Use Control-C to stop this server and shut down all kernels (twice to skip confirmation).

This indicates that the server is ready to accept connections on your local machine at port 8888. Here's a link for convenience: http://localhost:8888/. If port 8888 is not available (say if you're running more than one server, you may get a few more messages and the server may end up accepting connections on a different port. The messages in the terminal should tell you which port to try.

Once you open up the main notebook page, if you click on the notebooks/ directory, you should see the full collection of notebooks. Clicking on any of the notebooks will open them for running in an interactive python session. Clicking a cell and typing shift-enter will run the cell. The Jupyter Notebook documentation has more information on working with notebooks.