Data Science

Jupyter Notebook

Dentro del Virtual environment activado:

>pip install jupyterlab

>jupyter notebook

API is the file with instructions for calculations, the back end of a web app.

API:  fastAPI

Use when no parent folder:  uvicorn main:app --reload  # main is the name of file main.py with fastapi inside. 

When parent folder: uvicorn app.main:app --reload

then go to http://127.0.0.1:8000/docs         # there you see a web app with the input output of your script.

API - Web GUI:  fastAPI with Streamlit

Use streamlit. Create a file name_streamlit.py, with streamlit code inside, for web GUI.

>streamlit run  name_streamlit.py   # no se ejecuta como .py file.

Relational Database (SQL):  MySQL, PostgreSQL, Oracle, SQL Server

    SQLAlchemy (is an Object Relational Mapper) is the python library for SQL, PostgreSQL is the recommended.

No Relational Database (NoSQL):  MongoDB (document), Redis (key value), Neo4j (Graph), Apache Cassandra (wide column)

    MongoDB (is an Object Document Mapper) is the recommended. PyMongo (and MongoEngine) is the python library.

There are mainly 3 types of Cloud services:

SaaS (Software as a service) Gmail, Netflix;   PaaS (Platform as a service) Neon.tech, Heroku;   IaaS (Infrastructure as a service) AWS, Azure, Google cloud [PaaS uses one IaaS physical server to run on]

Neon tech is a cloud-native PostgreSQL database platform

MongoDB Atlas is the cloud service for MongoDB (NoSQL).