First of all, you will need Python in your machine and I higly recommend you to install an enviroment package like Conda.
A good summary of the most important commands to learn with Codna are in its cheatsheet.
- Basic Tutorial
conda create --name tutorial python=3.8;
conda activate tutorial;
conda install -c conda-forge -y pandas jupyter;
conda install -y matplotlib;
conda install -c conda-forge -y jupyter_contrib_nbextensions;
conda install -y seaborn;
conda install -c conda-forge -y dataprep;
conda install -c conda-forge -y pandas-profiling;
conda install -y scikit-learn;
- Adding TSFresh to your basic tutorial env
conda create --clone tutorial --name tutorial-tsfresh
conda activate tutorial-tsfresh;
conda uninstall -y pandas-profiling; # Try not creating this repo without removing this package;
conda install -y tsfresh;
conda install -y statsmodels==0.12.1;
- Adding Pycaret to your basic tutorial env
conda create --clone tutorial-tsfresh --name tutorial-pycaret;
conda activate tutorial-pycaret;
pip install --pre pycaret;
- Adding Keras to your basic tutorial env
conda create --clone tutorial --name tutorial-keras;
conda activate tutorial-keras;
conda install -c apple tensorflow-deps; # if you are using a mac;
pip install tensorflow-macos;