Part I · Quantitative Foundations
Module 01. Modern Data Engineering: Tidy Data, Polars & DuckDB
Turn raw agency feeds into a tidy, audited panel, the table every causal estimator runs on.
Macroeconomics: joining inflation and unemployment panels about 6 hours3 lessons
Module progress0/3 lessons · 0%
Lessons
1.1
Colab, the Modern Stack and Tidy Data
- · Explain why data architecture decides whether a causal estimate can be trusted
- · Set up a reproducible Google Colab notebook with Polars and DuckDB
- · Recognise wide, long and tidy layouts and convert between them with Polars
120 min 5 Colab cellsVideo coming soon
1.2
Relational Joins and SQL over DataFrames with DuckDB
- · Join tidy tables with DuckDB SQL directly on Polars DataFrames
- · Choose between INNER, LEFT and ANTI joins and audit what each one drops
- · Detect missing periods before they bias a panel estimate
120 min 6 Colab cellsVideo coming soon
1.3
Aggregation, Windows and the Analysis-Ready Panel
- · Compute grouped summaries and rolling statistics with Polars expressions
- · Build pre- and post-period baselines, the raw material of a before/after comparison
- · Use Polars lazy queries and save an analysis-ready panel to Parquet
120 min 7 Colab cellsVideo coming soon