Python Data Visualization («PYTDAT»)

• Preparing and analysing data
• Choosing the right visualization for the respective use case
• Getting to know the range of Python visualizations
• Creating professional and understandable diagrams
• Developing interactive visualizations
• Best practices for good visual storytelling

In this course, you will learn how to visualize data with Python in an impressive and understandable way.

From the basics with Pandas and Matplotlib to elegant statistical plots with Seaborn and interactive dashboards with Plotly - you will learn the most important tools of modern data visualization step by step.

Discover best practices, create your own visualizations and take your analyses to a new level - in a practical way so that what you learn can be applied directly in everyday life.

1 Refresh Data analysis with NumPy and Pandas

• Overview: Aim and sequence of the course
• Data analysis with NumPy
  • Arrays, data types, indexing
  • Relevance for numerical data visualization
• Level-up with Pandas
  • Series and DataFrames
  • Importing data (CSV, Excel)
  • Applied data exploration (head, describe, info, isnull, value_counts)
  • Simple corrections
• Mini exercise: First analysis of a real data set (e.g. Titanic, Iris)

2 Basics of visualization with Matplotlib

• Philosophy of Matplotlib
• Plot types:
  • Line, scatter, bar, histogram, boxplot
• Axes, titles, legends, colors, styles
• Subplots and layouts
• Saving plots (PNG, PDF, SVG)
• Mini exercise: Creating different plots with real data

3 Data analysis and visualization with Seaborn

• Why Seaborn? High-level vs. low-level APIs
• Plot types:
  • countplot, boxplot, violinplot, histplot, scatterplot, pairplot, heatmap
• Context analysis:
  • Visualizing correlations (heatmap)
  • Visualizing relational and categorical data
• Design and style customization
• Mini exercise: Exploratory data analysis with Seaborn

4 Visual storytelling and best practices

• Difference between exploratory and explanatory visualization
• Visual principles:
  • Colors, axes, scaling, annotations
• Common pitfalls: what makes a "bad" graphic?
• Introduction to design guidelines (data-ink ratio, Tufte principles)
• Mini exercise: Improving bad vs. good visualization

5 Interactive visualization with Plotly

• Why interactive visualization?
• Introduction to Plotly Express
  • px.scatter, px.bar, px.line, px.histogram, px.box
• Interactive features: Zoom, hover, tooltips
• Colors, facets, animations
• Mini exercise: appealing and interactive dashboards with Plotly Express

6 Plotly Graph Objects & Dash introduction

• Plotly Graph Objects (GO) vs. Express
  • Flexibility and customization
• Introduction to Dash (conceptual)
  • Building a simple dashboard
• Layouts and callbacks (demo)
• Mini exercise: Visualize a small interactive report

7 Other tools & special visualizations

• Altair: declarative visualization
• Bokeh: interactive web plots
• Geopandas & Folium: maps and geographic data
• Wordclouds & networks
• When to use which tool? Overview & comparison table

8 Mini project

• Mini project (individually or in small groups):
  • Analyze, visualize and present data
  • Use at least 2 libraries
• Goal: Communicate findings from data visually
• Short presentations & feedback round
• Conclusion: Q&A, tips for further learning

This course is characterized by a mixture of frontal teaching, guided exercises and examples as well as hands-on scenarios, adapted, extended or created by the participants. Each chapter concludes with a mini exercise.
The course ends with a project, which is designed to be exploratory and interactive.

This course is aimed at anyone who not only wants to analyze data, but also visualize it in a convincing way:

• Data analysts and data scientists
• Students and researchers from all disciplines
• Business analysts and controllers
• Developers with an interest in data visualization
• Anyone who works with data and wants to communicate their results in an understandable way

We recommend attending the following course in preparation:

The following must be installed by the participants in advance:

• Python (at least 3.10+)
• Python IDE (PyCharm or VSCode, etc.)
• Sufficient host rights to install packages with pip