vitoriastavis / ml-piotr

Machine Learning

Materials for the Machine Learning course.

Setting up the python environment

In this course you will be working with python using jupyter notebooks or jupyter lab (prefered). So first you have to set up a proper python environment. I strongly encourage you to use some form of a virtual environment. I recommend the Anaconda or its smaller subset miniconda. Personally I recommend using mambaforge as conda tends to be rather slow. After installing mambaforge create a new virtual environment ml (or any other name you want):

conda create -n ml python=3.9

Then activate the environment by running

conda activate ml

To close environment you type

conda deactivate

Now you can install required packages (if you are using Anaconda some maybe already installed):

mamba install  jupyterlab jupytext  ipywidgets
mamba install numpy scipy  scikit-learn
mamba install matplotlib

If you didn't install mamba then you can substitute conda -c conda-forge(-c conda-forge tells to add conda-forge channel which is turned on by default in mambaforge ) for mamba. I tend to use mamba as it is markedly faster then conda.

After installing all required packages you can start jupyter lab by running

jypyter lab

Rmd format

The notebooks in the repository are stored in Rmd (R Markdown) format. Thanks to the jupytext package you can open them right in the jupyter lab, by clicking the file name with righthand mouse button and choosing open with and then Notebook. If you are using jupyter notebook the you have to convert them prior to opening by running

jupytext --to notebook <Rmd file name>

Using python in lab

When using the computers in lab, please log to your linux account and then run

source /app/Python/3.10.4/VE/defaults/bin/activate

Then you can run

jupyter lab