Jupyter Notebook Data Science Workflow Training Course

Course overview

Data science teams often lose time moving between exploratory analysis, data cleaning, visualisation, model experiments and stakeholder reporting. Jupyter Notebook can bring these activities into one auditable working document, but only when notebooks are structured, reproducible and suitable for collaboration. This course addresses the gap between writing isolated Python cells and operating a dependable notebook-based workflow that colleagues can rerun, review and use to make decisions.

Participants build a disciplined workflow in Jupyter Notebook and JupyterLab for importing and validating data, transforming datasets with pandas, exploring distributions, creating decision-ready charts and training baseline machine-learning models with scikit-learn. The programme covers notebook architecture, kernel and environment management, markdown documentation, reusable functions, parameterised analysis, debugging, version control practices and export options. Participants learn how to distinguish exploratory work from production-ready analytical assets and how to make assumptions, data lineage and results visible.

Instructor-led demonstrations are followed by guided labs using a realistic business dataset with missing values, inconsistent fields and measurable decision questions. Each participant develops an end-to-end analytical notebook: a documented, reproducible analysis containing data-quality checks, transformations, visualisations, model evaluation and an executive-ready findings section. Peer review and instructor feedback help participants improve clarity, robustness and maintainability before they leave.

The course is designed for analysts, data scientists, developers and technically capable business specialists who already use Python or need to work confidently with Python-based data science teams. It is particularly valuable where notebooks must be shared, reviewed or retained as evidence for analytical decisions.

Course objectives

By the end of this course, participants will be able to:

• Configure Jupyter Notebook and JupyterLab kernels, extensions and project environments for repeatable analysis
• Structure analytical notebooks with markdown narratives, clear cell sequencing and reusable Python functions
• Import, profile and validate CSV, Excel and database extracts using pandas
• Clean and transform tabular data with joins, grouping, reshaping, missing-value treatment and date handling
• Create diagnostic and stakeholder-ready visualisations with matplotlib and seaborn
• Train and evaluate baseline scikit-learn classification and regression models using appropriate metrics
• Apply Git-based version-control practices to notebook projects while managing outputs and data files
• Produce a documented end-to-end Jupyter notebook that communicates methods, findings, limitations and next actions

Benefits of attending

Target competencies

Who should attend

• Data Analysts — who need to turn raw extracts into traceable analyses and decision-ready reports
• Data Scientists — who need more reproducible, reviewable notebook workflows for experimentation and model development
• Business Intelligence Analysts — who want to extend dashboard and SQL work with Python-based exploratory analysis
• Machine Learning Engineers — who need to collaborate effectively with analysts working in notebooks before deployment
• Python Developers — who need to apply existing coding skills to data preparation, visualisation and model evaluation
• Technical Product Managers — who need to assess notebook outputs, assumptions and model evidence from data teams

Requirements and prerequisites

Participants should be comfortable writing and running basic Python code, including variables, lists, dictionaries, loops, functions and imports. They should understand common tabular-data concepts such as rows, columns, data types, filters and joins, and should have used spreadsheets, SQL or a reporting tool to inspect data. Prior exposure to pandas is helpful but not essential. Participants need access to a laptop on which approved software can be installed or a browser-based training environment can be used. Prior machine-learning experience, advanced statistics, Git expertise and production deployment experience are not required.

Training methodology

The course combines short instructor-led demonstrations with frequent hands-on work in Jupyter Notebook and JupyterLab. Participants progressively analyse a realistic business dataset, beginning with data profiling and ending with model evaluation and an executive findings section. Labs focus on writing, running and debugging cells rather than watching demonstrations. Small-group code reviews examine notebook structure, visualisation choices and assumptions, while case discussions connect metrics to business decisions. On day five, participants refine their own end-to-end notebook and create an application plan for introducing the workflow in their team.

Course outline

Day 1: Notebook foundations and reproducible project setup

• Jupyter Notebook and JupyterLab interfaces, cells, kernels and command modes
• Python environments, package installation and kernel selection
• Notebook project folders, naming conventions and data-path management
• Markdown documentation, headings, equations, tables and links
• Code-cell execution order, hidden state and restart-and-run-all discipline
• Python functions, imports and configuration variables for reusable analysis
• Debugging notebook errors with tracebacks, inspection and assertions

Workshop: Participants create a structured notebook project and produce a documented data-ingestion notebook that can be restarted and rerun without manual fixes.

Day 2: Data acquisition, profiling and transformation with pandas

• Loading CSV, Excel and JSON data into pandas DataFrames
• Data types, indexes, schema inspection and memory-aware loading
• Missing-value profiling, duplicate detection and range validation
• Filtering, sorting and boolean masks for targeted analysis
• Grouping, aggregation and pivot tables for business measures
• Merging, concatenating and reconciling related datasets
• Reshaping, string cleaning and datetime transformation techniques

Workshop: Participants clean and combine customer, transaction and product files, producing a validated analytical dataset and a recorded data-quality summary.

Day 3: Exploratory analysis and visual communication

• Exploratory data analysis questions and hypothesis-led notebook sections
• Descriptive statistics, distributions and outlier investigation
• Correlation analysis and confounding considerations
• Matplotlib figure construction, axes control and annotation
• Seaborn categorical, distribution and relationship plots
• Chart selection for operational, customer and financial decisions
• Interpreting patterns, uncertainty and limitations without overstating conclusions

Workshop: Participants build an exploratory analysis section with four annotated charts and a concise written interpretation of the drivers behind a selected business outcome.

Day 4: Machine-learning experiments and notebook collaboration

• Problem framing for classification and regression use cases
• Train-test splits, leakage prevention and feature preparation
• scikit-learn pipelines for preprocessing and model fitting
• Baseline models using linear regression, logistic regression and decision trees
• Model metrics including MAE, RMSE, accuracy, precision, recall and ROC-AUC
• Git repositories, notebook diffs and output-clearing practices
• Parameterisation, reusable modules and criteria for moving beyond a notebook

Workshop: Participants train and compare two baseline models in a reusable notebook, then commit a cleaned version to a Git repository with a meaningful README.

Day 5: Publishing findings and applying the workflow at work

• Notebook narrative flow from business question to recommendation
• Documenting assumptions, source lineage and analytical limitations
• Executive summaries, key findings and action-oriented recommendations
• Notebook export to HTML, PDF and presentation-ready formats
• Peer review checklist for code, data, visuals and conclusions
• Notebook governance, access control and sensitive-data handling
• Personal workflow adoption plan for team standards and review routines

Workshop: Participants complete and present an end-to-end analytical notebook, receiving peer and instructor review before producing a practical implementation plan for their workplace.

Tools & standards covered

Jupyter Notebook, JupyterLab, pandas, scikit-learn

A typical training day

Live online deliveries follow the same structure in the East Africa Time zone, with shorter screen blocks and longer breaks.

What the fee includes

How you can take this course

Certification

Participants who complete the full five days receive the Skillset Development Certificate of Completion, stating the course title, course code, dates and delivery format — suitable for professional-development records and employer reimbursement.

Frequently asked questions