Streamlit: The fastest way to create and share data applications!

Transform data scripts into shareable web applications in minutes, all in Python, and for free with no front-end experience required.

Installation:

It is necessary to have Python 3.6 or higher, and the installation will be done via PIP:

pip install streamlit
streamlit hello

Right after that, it is necessary to configure some environment variables, depending on the operating system you are using:

Streamlit is compatible with many libraries and frameworks such as Keras, Scikit learn, Altair, bokeh, Latex, plotly, OpenCV, Vega-Lite, PyTorch, NumPy, seaborn, Deck.GL Tensorflow, matplotlib, pandas, and many others

Components:

With the popularization of Streamlit, several components appeared with the most diverse functionalities, such as:

HiPlot

A lightweight interactive visualization tool to help AI researchers discover correlations and patterns in high-dimensional data using parallel plots and other graphical ways of representing information.

pip install hiplot

Gallery:

There are many examples of its use and configurations of different apps; thinking about demonstrating its usability by the community, a gallery was created with several demonstrations of its use.

Happy example :)

This demo project below allows you to browse the entire Udacity autonomous car dataset and perform inference in real-time using the YOLO object detection network.

Deploy Apps:

One of its great differentials is the possibility of deploying apps in a very automated way using GitOps, that is, through continuous integration, the developer, from a given push on a branch in a code repository like GitHub, triggers an automation that does to deploy the app.

This framework is straightforward and easy to use, the deployment of new apps is very intuitive. As a negative point, I missed a more visible layer of security, and I think I will have to implement an interface between the user and the apps to ensure more security….

References: