Setting up Julia Environment (Mac/OS)

Marc Brittain

1. Download Julia version 0.5

Make sure to download a copy of Julia 0.5 and not the newer version: Julia 0.6. Some of the modules are not yet compatible with 0.6
Follow this link to download a copy of Julia 0.5: Julia 0.5
Install like any other program and you should be able to run basic commands in the Julia interpreter that appears once the installation finishes

2. Jupyter Notebooks

Next it is time to install the Jupyter notebook extension to Julia. The recommended way to install Jupyter notebooks is to first install Anaconda. Installing Anaconda will also install a copy of Python onto your device. Although you may not explicitly use python, Julia can use different python functions so it is important to have it installed
Follow this link to download a copy of Anaconda (Choose python 3.6 option): Anaconda
- [Note]: If you already have Anaconda installed with python 2.7 then you do not need to install the newer Anaconda version, the best route is to create a new python 3.6 environment. See Supplemental Notes at the bottom of the page if this applies to you
After you have installed Anaconda, it is time to make Jupyter compatible with Julia. To do this, navigate to Julia through either the menu or desktop icon (This should open the Julia interpreter)
- In the interpreter type the following:
```
                    Pkg.add("IJulia")
```
Once the above command is finished, open up a new terminal and type:
```
                    jupyter notebook
```
- Jupyter is installed properly when the above command opens the Jupyter application and displays the different directories on your device
  - [Note]: I advise creating a new folder to store julia notebooks and if you work in python create a different folder for python. This will help down the road for navigating and organizing work
    - To do this, locate New in the top right of the screen and select Folder. Once you select the folder rename it to something related to julia or to the course

3. Download course code

Download and unzip the aere504-notebook directory: aere504-notebook-archive.zip from the following link: GitHub
Once you have downloaded the course code, move that folder into the julia folder that was created in step 2. This can be done in Finder or in Jupyter by opening the new folder and selecting upload
To install all required packages, create a julia notebook by selecting New and then Julia
In the first cell block, type the following code and then hit shift-enter, or shift-return to execute the cell:
```
include("aero504-notebook-master/install.jl")
```
This will take some time to run but will install the modules needed for the course code

4. Installing Homebrew

We now need to install a copy of homebrew on the device. To do this follow the link: Homebrew

5. Installing final package dependencies

The final packages to install are: pdf2svg and MacTeX
- pdf2svg
  - In terminal type:
```
        brew install pdf2svg
```
- MacTeX
  - MacTeX can be downloaded here MacTeX
  - Note: MacTeX may take some time to download

Once both packages are installed, restart your device

6. Fixing Bayes Net script

One of the modules will encounter an error due to code that is outdated in the script file. This is easy to change once you find the file.
To find the file, open up a new terminal window and type the following command:

                open ~/.julia/v0.5/BayesNets/src

Right-click on the io.jl file and open with TextEdit or an application that can open code files
On line 9 you will see the code:

                plot(Graph(1), ["Empty Graph"])

This needs to be changed to the following:

                plot(DiGraph(1), ["Empty Graph"])

Once it is changed, save the file and Julia should now be configured to run all the course code!
You can also delete the notebook that was created to install the required packages, that is no longer needed.

Supplemental Notes

a. Creating python 3.6 environment

This section addresses creating a new python 3.6 environment if you already have Anaconda installed with python 2.

What is a python environment?
- A python environment is what you are working in when you start Anaconda for the first time. This is known as your root environment and defaults to the python version you installed with Anaconda. What happens if you are working on a project and python 2.7 and also want to start a new project in python 3.6? Creating a new environment solves this issue without having to reinstall Anaconda. In practice it is recommended to create a new environment each time you start a new project so that different modules don't conflict with each other.

To create a new python environment, run the following code in the command line and follow the instructions:

            conda create -n my_env_name_here python=3.6

Note: You create your own environment name

To activate your new environment run:

            source activate my_env_name_here

Then run the jupyter notebook command to start a python notebook in your new environment