WeAreDevelopers LIVE • Feb 17, 2021

Python Data Visualization @ Deepnote (w/ PyViz overview)

Radovan Kavický

Stop letting overplotting obscure critical signals in your massive datasets. Master the PyViz ecosystem to build scalable, AI-ready visualizations that instantly establish stakeholder trust.

Pause
Mute Enter Fullscreen
#1 about 4 min

Overview of presentation structure and interactive Slido setup

The session introduces the agenda covering Python history, visualization tools, common pitfalls, and machine learning trends.

#2 about 6 min

Speaker background and audience polling on Python usage

Community involvement context is established alongside a live audience poll determining the ratio of developers to data scientists.

#3 about 4 min

History and original philosophy of the Python programming language

Python was initially designed as an accessible educational language before evolving into a core tool for computational science.

#4 about 7 min

Ecosystem overview of Python data visualization libraries and tools

The visualization landscape spans Matplotlib-based libraries, JavaScript interfaces, and high-performance renderers that support diverse data analysis applications.

#5 about 5 min

Historical foundations of exploratory data analysis and visual grammar

Foundational work by John Tukey and Leland Wilkinson established the theoretical models underlying modern visualization code.

#6 about 10 min

Live Matplotlib rendering and data plotting within Deepnote environments

Interactive Python notebooks execute scripts to demonstrate plotting coordinate loops and overlapping data arrays using alpha transparency.

#7 about 5 min

Development history of scientific computation libraries and PyViz tools

Numeric processing backends merged into accessible open-source libraries that currently define the overarching PyViz ecosystem.

#8 about 8 min

Resolving overplotting and color saturation issues in dense visualizations

Techniques such as structural jittering, point size reduction, and dynamic data shading prevent large datasets from obscuring graphical insights.

#9 about 6 min

Automated visualization trends merging machine learning and business intelligence

Future workflows will rely on natural language queries and automated generative modeling to extract algorithmic insights from raw datasets.

Matching moments

5:36 min

Building a data analysis stack with Python and Jupyter

Markus Harrer Markus Harrer · World Congress 2021

1:48 min

Leveraging Python visualization libraries for hardware accelerated rendering

Prof Smoke Prof Smoke · World Congress 2025

1:50 min

Introduction to the speaker and data science background

Bas Geerdink · LIVE

3:33 min

Refactoring data science workflows using Rapids QDF and Pandas

Paul Graham Paul Graham · LIVE

2:11 min

Exploring interpretability and future advancements in tabular processing

Adrian Spataru +1 · LIVE

2:38 min

Overview of Databricks and interactive data processing

Alan Mazankiewicz · LIVE