Generate Professional Jupyter Notebook Templates

This skill creates well-structured, professional Jupyter notebook templates for data science, machine learning, and research projects. Every notebook follows a fixed nine-section flow: header and metadata, setup and configuration, data loading and overview, exploratory data analysis, data processing, modeling or analysis, results and evaluation, conclusions, and references. Cell organization follows fixed conventions - markdown cells for section headers and explanations, code cells scoped to a single task, clear variable naming with inline comments, and cell tags for organization and automation.

Use it when starting a new data science or research notebook that needs to be reproducible and shareable, not a quick one-off scratch analysis. It is not meant to ship with exploratory clutter: the export and sharing pattern strips all cell outputs before distribution, and the header template requires an explicit objective, dataset description, key questions, and expected outcomes up front rather than a title alone.

Given a new analysis, it produces a professional header template (title, author, date, version, and environment metadata, followed by objective, dataset, key questions, and expected outcomes), a standard setup cell (the core pandas, numpy, matplotlib, seaborn, scipy, and scikit-learn imports, a fixed plot style and random seed of 42, path and figure-size constants, and Jupyter display configuration), a data-loading function with error handling that validates expected columns and reports shape, memory usage, and a missing-value summary table, an EDA section with a reusable subplot-figure helper for numerical distribution histograms and categorical value-count breakdowns, and a results documentation template covering a data-quality assessment, numbered findings with supporting evidence, a statistical summary table, conclusions with prioritized recommendations, a next-steps checklist, and a references section for data sources and documentation links.

### Clean notebook for sharing
from nbconvert.preprocessors import ClearOutputPreprocessor
from nbconvert import NotebookExporter

def clean_notebook_output(notebook_path):
    """Remove all outputs from notebook for clean sharing"""
    clear_output = ClearOutputPreprocessor()
    exporter = NotebookExporter(preprocessors=[clear_output])
    
    with open(notebook_path) as f:
        nb = f.read()
    
    (body, resources) = exporter.from_filename(notebook_path)
    
    with open(notebook_path.replace('.ipynb', '_clean.ipynb'), 'w') as f:
        f.write(body)

Advanced features add version-control provenance (capturing the current git hash and branch via subprocess calls, plus an execution timestamp) and performance monitoring (a decorator that wraps any function to log execution time and memory-usage delta via psutil), so results in the notebook can be traced back to the exact code state and cost that produced them.

Data scientists, ML engineers, and researchers who need a consistent, shareable notebook structure across projects, with domain-specific adaptations - model comparison and hyperparameter tuning sections for ML projects, KPI tracking and metric definitions for business analytics, methodology and statistical test documentation for research, and seasonality and forecasting sections for time series work.