Dtale Tutorial – Guide To Visualize Pandas Data Structure

Analyzing a data structure helps us gather information about the data like how it is stored, what are the different attributes and their properties. Data Analysis can be performed using different python libraries like pandas, etc.

DTale is a Flask and React-based powerful tool which is used to analyze and visualize pandas data structure seamlessly. It supports different objects like Data Frame, Series, etc. It works beautifully on both the Jupyter notebook and the command-line interface. 

DTale is a Graphical Interface where we can select the data we want to analyze and how to analyze using different graphs and plots. 

In this article, we will explore Dtale with all its functionalities.

Implementation of Dtale

Like any other python library, we need to install DTale before using it by pip install dtale.

Download our Mobile App

1. Importing required libraries

We will import Dtale before using it, also we will be downloading our dataset from plotly so we also need to import plotly. You can use any dataset you have.

import dtale

Import plotly.express as px 

import pandas as pd

2. Loading the dataset

For this article, we will be downloading the dataset named ‘tips’ using Plotly. The dataset contains different attributes of restaurant data. 

df = px.data.tips()

df.head()

3. Loading Dtale

d = dtale.show(df)

#open it in a new window in browser

d.open_browser()

This is the data frame visualized using Dtale. Here we can change the value of any attribute to our desired value. In this image, you can see the highlighted play button.

When we click on this button we see different functions that are displayed in the image below.

Here we can see that it provides us with different options like:

1. Describe: Used to display the statistical properties of the dataset.
2. Custom Filter: To apply filters to columns
3. Build Columns: Used to build new attributes and add values to it.
4. Visualization: By using Correlations, Charts, Heatmaps we can visualize the dataset.
5. Export: To export the CSV with changes that are implemented.

These different functions are used to analyze the dataset and how we want to want to manipulate the data according to our demands. We can visualize the dataset by adding new attributes to our dataset and also export the dataset with the required changes.

Now let us analyze the column-wise functions and see what all operations we can perform on the columns of the dataset.

By clicking on the column name we can see a dropdown menu as shown in the image above. The functions defined here are:

1. Arranging the columns in Ascending or Descending Order.
2. Lock or Hide the column
3. Delete the column and rename the Column.
4. Describe statistical analysis of the column.  Etc. 

We can click on different options/functions to see how it works.

1. Describe 

Describe function is used to display the statistical properties of the dataset. Lets us analyze the statistical properties of this dataset by clicking the describe option.

Here we can see that we can analyze the statistical properties of different attributes by selecting them using the checkbox provided. Also, you can visualize the histogram of the data by clicking on the Histogram option. 

Not only this you can also copy the code which is running in the backend to show the current visualization by clicking on the code export option. 

2. Visualization

• Correlations

Here we can plot correlations between different attributes. We can select these attributes using widgets. It not only allows you to visualize but also displays the value of the correlation.

2. Charts

Charts help us visualize all different kinds of charts between different attributes. It also supports 3d Charts and animations with the group by and aggregation functions also. It is really useful as you can visualize any type of relationship using different charts.

3. Manipulating Data

Other functions like Highlight Range, Highlight dtypes, Highlight Missing, etc. are used to manipulate the data according to our wish and what we want to visualize or analyze. It helps us take out the outliers, ambiguous values, any irregularity in the data, etc. 

Dtale helps us to do an Exploratory Data Analysis of the dataset and that too without even writing more than one line of code. It is a GUI Based Exploratory Data Analysis tool that analyzes and visualizes each and every aspect of the dataset. It is easy to use and blazingly fast.

Dtale can be used to create detailed EDA reports and that too with minimum effort and maximum output. If we go by the conventional way of creating the EDA report it will take a lot of time while with Dtale we can create it in an hour.

Conclusion

In this, we learned about Dtale: EDA Reporting tool which is highly useful and is used to analyze and visualize all types of datasets. We saw how Dtale can save our time and efforts to create visually appealing and insightful EDA reports.