Python for Data Analysis and Visualization

Set up a working environment and learn the Python constructs used in everyday analysis.

• Jupyter notebooks, cells, and Markdown documentation
• Variables, strings, numbers, and f-strings
• Lists, dictionaries, and tuples
• Loops, conditionals, and list comprehensions
• Writing and reusing functions
• Installing packages with pip and managing virtual environments

Hands-on lab: Create a notebook that reads a small list of records, computes summary figures with functions, and documents each step in Markdown.

Work with arrays and DataFrames, the core data structures behind Python analysis.

• NumPy arrays, vectorized operations, and broadcasting
• Series and DataFrame structure
• Reading CSV, Excel, and JSON files
• Selecting with loc, iloc, and boolean filters
• Sorting, ranking, and basic descriptive statistics
• Reading data from SQL with pandas

Hands-on lab: Load a public sales dataset from two file formats, filter and rank records, and replace a spreadsheet summary with a few lines of pandas.

Turn messy raw data into a trustworthy table, and record every change you make.

• Detecting and handling missing values
• Removing duplicates and fixing data types
• String cleaning with the str accessor and regular expressions
• Parsing dates and working with time zones
• Detecting outliers with simple rules
• Building a repeatable cleaning function

Hands-on lab: Clean a deliberately messy customer export, document each decision in the notebook, and produce a validated, analysis-ready file.

Reshape and join tables to answer questions that span multiple sources.

• groupby with aggregation and named aggregations
• Merge and join types and checking for row explosion
• Pivot tables and melt for wide and long formats
• Creating derived columns and binning
• Rolling and resampled time-series summaries
• Method chaining for readable pipelines

Hands-on lab: Join orders, customers, and products tables, then compute monthly revenue by segment using a single readable pandas pipeline.

Explore a new dataset methodically to find structure, quality issues, and questions worth answering.

• Profiling shape, types, and distributions
• Univariate and bivariate analysis
• Correlation and its limits
• Segment comparisons and cohort views
• Spotting Simpson's paradox and other misleading aggregates
• Writing down hypotheses before testing them

Hands-on lab: Run a structured exploratory analysis on an unfamiliar dataset and write a one-page list of findings, caveats, and follow-up questions.

Choose the right chart for the question and build clear, accurate static visuals.

• Matching chart types to questions
• matplotlib figures, axes, and styling
• seaborn distribution, categorical, and relational plots
• Small multiples and faceting
• Color, labeling, and accessibility choices
• Avoiding misleading axes and cluttered charts

Hands-on lab: Redraw three poor-quality charts from a supplied report into clear, labeled, accessible versions and explain each design change.

Build interactive charts with Plotly and shape an analysis into a narrative for a decision-maker.

• Plotly Express charts, hover details, and filters
• Combining multiple charts in a dashboard-style layout
• Structuring a story: context, evidence, recommendation
• Writing chart titles that state the takeaway
• Exporting notebooks to HTML for sharing
• Documenting assumptions and data limitations

Hands-on lab: Convert your earlier analysis into an interactive notebook with three Plotly charts and a short written recommendation, then present it to the cohort.