datamade/data-analysis-guidelines

📒 Analyzing Data, the DataMade Way

Analyzing Data, the DataMade Way

⚠️ Deprecation warning: This documentation no longer represents DataMade's current best practices for data analysis. For contemporary guidance on data analysis, refer to the how-to repo.

You've extracted and transformed the data. Now it's time to load (analyze) it. Here, you'll find the principles that inform DataMade's approach to data analysis, as well as the tools and organizational practices that make it possible.

Principles

DataMade's approach to data analysis combines our principles for making data with the basic principles of literate programming.

Namely, data analysis should:

1. be reproducible with one command.
2. be conducted using standard tools.
3. be kept under version control.
4. prioritize legibility to other humans.

Guides

• Data analysis 001 - Setup
  • Setting up your environment
  • Organizing your analysis
• Data analysis 101 - Standard toolkit
  • Introduction to pweave
• Data analysis 201 - Putting it all together
  • Patterns for multi-part analysis
• Appendix A - LaTeX
• Appendix B - pandas
• Appendix C (external issue) – matplotlib

Examples

Under construction in the examples dir! 👷