Machine Learning, Exploratory Data Analysis

One-Line, Magical Code to Perform EDA!

One line solves all your problems!

Daksh Trehan

Published in

Towards AI

5 min readApr 24, 2021

“Data is the new oil” ~ Clive Humby

Data is an integral part of our life and unlike other resources it is inexhaustible but here comes a catch, it is only useful to your organization if you know how to mend it and get its gist.

Data Science is a process that includes: Collecting, Storing, Processing, Describing, and Modeling.

Processing, Describing/EDA(Exploratory Data Analysis) can be referred to as a lifecycle to get introduced to the data by finding relations among each variable and visualizing them to find hidden trends.

EDA accounts for a large amount of time and effort to clean and explore our data. Though in the field of data processing we still expect some advancements for Data Exploration there have been astounding improvements. Several open-source libraries have come up with a no-code or low-code method to help ease the exploration.

D-Tale is one such library, it is the combination of Flask back-end and React Front-end that brings up an interactive way to visualize and explore pandas data frame.

D-Tale makes sure you aren’t sick of performing df.head() recursively!

Implementing D-Tale

Install D-Tale: Like any other Python Library you can easily install D-Tale by using “pip install dtale” in your command line.

2. Importing Relevant Libraries: Use Seaborn to load dataset and D-tale to visualize and explore.

3. Chose a dataset: From pre-defined dataset in Seaborn chose any one.

4. Take basic insights of data: Use describe() method for the same.

5. Use D-Tale: Load the data using D-Tale library and get interactive insights.

Tap on play button and chose “Open in New Tab”:

Features of D-Tale

All you Data in clean and elegant way!

Look for Statistical Property of your dataset using “Describe” button.

Summarize the data of relevant rows and columns.

Remove Duplicate Values

Check the Correlation in your data for easy creation of charts.

The library helps you with agnostic score that helps to determine linear and non-linear relationship.

The most anticipating feature of D-Tale is its ability to create charts seamlessly.

It offers a wide variety of charts with data cleaning options that are personalized to each chart type.

Another highlighting feature of D-Tale is that it allows you to create chart from the dashboard and directly import the code for the same.

Heat maps can also be employed either on whole data or on particular columns.

You can also check statistical values for particular column and perform column analysis.

Perform Variance analysis on each column.

The user can also change the data of any particular record.

The above mentioned code can be found at: dakshtrehan/D-Tale-Exploration (github.com)

If you like this article, please consider subscribing to my newsletter: Daksh Trehan’s Weekly Newsletter.

Conclusion

The article helped us to throw a light on extremely powerful EDA reporting tool: D-Tale. We saw how D-Tale make it super easy to create appealing visuals and explore the data.