3 Python Packages for Low Code Data Science

1. Mito

Mito is a spreadsheet interface for Python. Mito allows the users to perform data science actions like exploratory data analysis, visualizations, and reading in Excel files. Each edit you make in the Mito will generate the equivalent Python in the code cell below. The package gives users a visual environment to do their data analysis (the environment should be familiar since it is modeled after a spreadsheet) and generates all the code for the user, saving time on trips to Stack Overflow or Google.

From Author

Mito has features such as pivot tables:

From Author

And the ability to dive into summary statistics for your dataset:

From Author

Mito also offers:

• Merging
• Filtering
• Sorting
• Adding and Deleting Columns
• Saving and Replaying Analyses (macros)
• and more!

To install Mito, run these commands:

python -m pip install mitoinstaller
python -m mitoinstaller install

Then open Jupyter Lab and call the Mitosheet

import mitosheet
mitosheet.sheet()

Here are the full install instructions.

2. Streamlit

Streamlit is an amazing tool for making interactive data apps. Within Streamlit, you can spin up a user friendly UI that allows colleagues and external end users to engage with a dataset interactively. Instead of just a notebook that users can run through to see an analysis, users can interact with the analysis themselves, changing parameters and toggling visualizations to more clearly understand the author’s takeaways, as well as come up with their own.

It is becoming more and more important for data scientists and data science teams to effectively communicate their analyses outside of the data science vertical. Streamlit makes it possible for data science teams to host data apps that everyone from engineering to finance to marketing can engage with.

To start working with Streamlit’s open source framework, run these commands:

pip install streamlitstreamlit hello

https://streamlit.io/

The above image shows the interactive widgets you can make with Streamlit. The tool allows you to take Python script and convert specific pieces of it (or the whole thing) into these widgets that can be hosted on a web page for easy end user access.

Here is the full documentation.

3. Lux

Lux is the only package in this article that is focused on solely on charting. Instead of giving the user an environment to create their charts, Lux allows the users to pass in their DataFrame and then it will automatically generate and suggest charts that the user can pick from. This is an extremely fast way to get great charts made, especially if you are making very standard charts in a repeated fashion. The actual coding of visualization packages can really slow down a workflow. Lux is a great way to circumvent this.

https://github.com/lux-org/lux

To import Lux:

import lux
import pandas as pd

Leave a comment, if you have a chance to use any of these packages :)