Package Manager and Environments

[!info]- Common used python packages
- numPy for working with custom data types
- pandas for working with DataFrames.
- scikit-learn for machine learning.  - Matplotlib for visualizations.
- seaborn for statistical visualizations.
- NLTK for text analysis.
- spaCy for more text analysis.

There are several virtual environments handler. Here’s a nice thread giving brief about those manageres https://stackoverflow.com/questions/41573587/what-is-the-difference-between-venv-pyvenv-pyenv-virtualenv-virtualenvwrappe # venv I like to use the shipped manager venv because it’s simple to manage and doesn’t require install it additionally

python -m venv envName
cd envName/scripts 
activate.bat //Activates the virtual enviroment in cmd
source activate //Activates in bash
deactivate //deactivates in bash
where python //returns the path for python interpretor 
get-command -all python //Same as above but used in powershell
pip freeze >> requirements.txt// returns the dependencies of the workspace and save them to requirements.txt file
pip install -r requirements.txt

Q. What is conda, mamba, mambaforge, miniforge?

Conda is a package and environment manager similar to pip for python. Mamba is C++ based implementation of conda offering faster package installation MambaForge is mamba based distribution which contains a preset collection of set packages for quicker setup and ready to go operation feasibility. Miniforge is conda based distribution

pip does not manage the installation of any code other than python. Eg: PyTorch relies on CUDA so, PIP isn’t the preferable way of installing it instead use conda/mamba Any python library can be installed using PIP

conda - Annaconda

https://www.dataquest.io/blog/python-vs-anaconda/

conda --version
conda install jupyter
jupyter notebook