Software setup

All software for the course can be installed without having administrator privileges on your computer. We achieve this by using a package manager called Miniconda. Miniconda creates isolated environments where software packages and package dependencies can be installed from a small file listing the required packages.

Wifi instructions
UF users and Eduroam users should sign into the SSID “UF” or “Eduroam”. USDA users should choose the SSID “UF Guest” and open their browsers to http://www.ufl.edu. This will bring up the network sign-on screen. [Click here]((https://net-services.ufl.edu/provided-services/wireless/UF-guest-wireless.html)for more connection information.

Windows 10 instructions

(modified from Nicholas Dawson’s instructions)

Download

Go here to download the latest version of Miniconda. Download Miniconda for Python 3.7 (not2.7) Download either the 32-bit or 64-bit Windows version that corresponds to your computer’s architecture. If you don’t know which architecture you’re using, check this out. Most users will download the 64-bit version.

Anaconda download page

Install locally

Open the downloaded Miniconda installer and select the installation type Just Me (recommended).

Anaconda install page 1

Choose the install location

Choose the default install location if it is available. That should be:

C:\Users\{username}\AppData\Local\Continuum\miniconda3

Anaconda install page 2

Choose advanced options

Select the default advanced options.

Anaconda install page 3

Post install

Your Miniconda shell should now be available under the start menu

Anaconda install page 5

Conda install

On Tuesday morning we will set up our conda environment as a class.

Download the this conda environment file windows64-mlenv.yaml. The file specifies all the conda packages that need to be installed in an isolated conda environment named mlenv.

From the Windows start menu select /Anaconda Powershell Prompt (miniconda3)

Move into the directory with the file and run:

conda env create -f windows64-mlenv.yaml

For more information refer to the conda docs

Once your conda environment has been created activate the environment:

conda activate mlenv

Download the course materials

On the first day of the course download the course materials so you can interactively code during the course.

Unzip the file then move into the top level of the directory in your Anaconda Powershell with the environment mlenv active.

Start your Jupyter notebooks

Start up your Jupyter notebook with this command:

jupyter notebook

This should open in your browser

Jupyter home screen

From the web page select assets > nb-lessons > 01_Python_for_ML_warm_up.ipynb

MacOS 10.14 instructions

Download

Go here to download the latest version of Miniconda. Download for Python 3.7 (not 2.7). Download the Mac OS X 64-bit (.pkg installer)

Anaconda download page

Install locally

Open the downloaded Anaconda installer and install in the default location

Anaconda install page

Post install

You can now open up your mac’s Terminal program under Applications -> Utilities -> Terminal.app. If you see (base)at the beginning of your Terminal prompt then conda was installed successfully.

Terminal

On Tuesday morning we will set up our conda environment as a class

Set up your conda environment

Download the this conda environment file mlenv.yaml. The file specifies all the conda packages that need to be installed in an isolated conda environment names mlenv. Open the Terminal.app and run this conda command in the directory with the file:

conda env create -f mlenv.yaml

For more information refer to the conda docs

Once your conda environment has been created activate the environment using

conda activate mlenv

Download the course materials

On the first day of the course download the course materials so you can interactively code during the course.

Unzip the file then move into the directory.

Start your Jupyter notebook

Start up your Jupyter notebook with this command:

jupyter notebook

This should open in your browser

Jupyter home screen

From the web page select assets > nb-lessons > 01_Python_for_ML_warm_up.ipynb

This will bring up our first notebook lesson. Go to the menu Kernel > Change kernel and select the kernel for our conda environment mlenv.

Linux instructions

Download

Go here to download the latest version of Miniconda. Download for Python 3.7. Download either the 32-bit or 64-bit Linux version that corresponds to your computer’s architecture. If you don’t know which architecture you’re using type arch into the terminal. Download the Linux 64-bit (bash installer) or the Linux 64-bit (bash installer).

Anaconda download page

Install locally

Open the terminal and move into the directory where Miniconda was downloaded and run the Bash script to install Miniconda

bash Miniconda3-latest-Linux-x86_64.sh

Post install

Close and reopen your Terminal. If you see (base)at the beginning of your Terminal prompt then conda was installed successfully.

Set up your conda environment

On Tuesday morning we will set up our conda environment as a class.

conda create -n mlenv python=3.7 pandas numpy seaborn jupyter
conda activate mlenv
conda install -c https://conda.anaconda.org/biocore scikit-bio
conda install -c conda-forge scikit-learn

For more information refer to the conda docs

Once your conda environment has been created activate the environment using

conda activate mlenv

Download the course materials

On the first day of the course download the course materials so you can interactively code during the course.

Unzip the file then move into the directory.

Start your Jupyter notebook

Start up your Jupyter notebook with this command:

jupyter notebook

This should open in your browser

Jupyter home screen

From the web page select assets > nb-lessons > 01_Python_for_ML_warm_up.ipynb

This will bring up our first notebook lesson. Go to the menu Kernel > Change kernel and select the kernel for our conda environment mlenv.

About Jupyter notebooks

We will be doing all of our coding in Jupyter notebooks so it’s useful to understand what these documents are. Jupyter notebooks are a mixture of text, code blocks and the output of those code blocks. When you first open up a notebook it usually has output graphs already generated. These outputs are just the saved from the last time the notebook was used. Behind the scenes you are opening up a blank new python terminal. No variables, functions or modules exist yet.

As you run code blocks the variables will be created and output will be replaced. Because of this the order you run these block in matters. If you run the last block first the variables it needs may not be available. Similarly, if you go back and run the same block again you can potentially overwrite variables. If you are getting unexpected results a common solution is to restart and run the code block sequentially. Remember that as you work and change values your notebook changes.