Aleksandra Badaczewska

Aleksandra Badaczewska

Alex is a Research Scientist IV at the Genome Informatics Facility at Iowa State University. Her academic background is in Chemistry and Biotechnology, with a Ph.D. in Computational Biology and broad experience in programming and designing web applications. She develops a comprehensive collection of highly customizable visualization solutions for Bioinformatics and supports software optimization for the USDA Geospatial analyses.

DataScience Workbook / 08: Data Visualization / 2. Introduction to scientific graphing / 2.2. Plotly-Dash: interactive plotting with Python / 2.2.3. Plotly graphing - interactive examples in the JupyterLab

Introduction

Python Interactive Graphing
using Plotly-Dash library and Jupyter Lab development environment

This tutorial introduces practical examples of interactive graphing in Python using the Plotly library. Our in-depth guide covers the ready-made examples stored in the data_graphing repository, which includes two Jupyter notebooks for each example - RTU and DIY.

  • The RTU notebook is a ready-to-use web application built with the Dash framework. It comes with built-in options for loading inputs and adjusting graphs on the fly. Simply Run all cells in Jupyter Lab to access the user-friendly interface of the app and interact with the graph. Use this option if you need a graph from your data without developing the app yourself.

  • The DIY notebook, is a do-it-yourself code variant using only the Plotly library and Python engine. The graph adjustment options are encoded in Python variables, and the application is split into functional blocks of code with explanations in markdown comments. This example allows users to learn how to develop a customized app from scratch and gain a deeper understanding of app development. Choose this option if you want to learn how to add advanced options to the app.

Overview

Python’s Plotly and Dash libraries can be used to create interactive graphs and web applications for data visualization. The Plotly library is built on top of Plotly.js (JavaScript plotting library), and allows users to create on-the-fly a range of customizable graphs. Dash is a web app framework that allows users to build interactive web apps using Python engine and Dash HTML Components. Together, these tools can be used to create engaging and informative visualizations of data that facilitates in-depth analysis.

Jupyter Lab provides a modern and flexible web-based user interface that makes it easy write and execute Python code, as well as include markdown-rendered text (e.g., documentation, instructions, comments) and visualizations (e.g., interactive graphs, figures), all-in-one document. The Jupyter notebooks (.ipynb) are a popular way to share and collaborate on data analysis and scientific computing projects.

Before you start…

To create interactive graphs in Python, you will need to install some libraries. One way to do this is to install the libraries globally on your system. However, a better option is to use a virtual environment specifically for efficient graphing. This approach uses Conda environment manager to separate the installation of the required libraries from your generic settings in the operating system. This helps keep your system clean and organized.

Requirements

These libraries contains predefined functions that you can call on your own dataset to create customized graphs.

python=3.9                      # a programming language that provides the environment for the following libraries
pandas=1.4.0                    # to import data from file or url, then manage data structure via DataFrame object
plotly=5.6.0                    # a library for creating customized interactive graphs
plotly_express=0.4.1            # a high-level wrapper for Plotly (an alternative approach to Graph Objects)
dash=2.1.0                      # (optional) to add on-the-fly customization of the graphs via widgets
dash_bio=1.0.1                  # (optional) to have predefined types of traces for specific biology-related tasks
kaleido                         # to export static images to any format
jupyter                         # to have Jupyter Notebook (file menagement in the separate tab)
jupyterlab>=3                   # to have Jupyter Lab      (next-generation user interface)
ipywidgets>=7.6                 # to display interactive graphs directly in the notebook
jupyter-dash                    # to add widgets (sliders, buttons) to Plotly charts in JupyterLab

Set up Conda environment

If you don’t have a virtual environment manager for your Python programming, consider using Conda. Simply follow the installation instructions in the Install developer libraries / Conda tutorial to get started. Once you have Conda installed, create a new virtual environment for interactive graphing.

  • Create virtual environment 
    conda create -n data_graphing python==3.9
  • Install Requirements 
    pip install pandas==1.4.0
pip install plotly==5.6.0
pip install plotly_express==0.4.1
pip install dash==2.1.0
pip install dash_bio==1.0.1
pip install -U kaleido
pip install jupyter
pip install "jupyterlab>=3" "ipywidgets>=7.6"
pip install jupyter-dash
  • Activate data_graphing environment 
    conda activate data_graphing

Launch Jupyter Lab in the terminal window

jupyter lab

This will open a browser on a localhost, by default http://127.0.0.1:8888.

On the left-hand side you can browse your local file system to enter the desired working directory. Then, click twice on the selected notebook (any .ipynb file) to open it in the right-hand side panel.

Below, you can learn basics, which allow you to run and customize with a success all pre-implemented examples available in the repository.