Data Science — Pandas Practice for Central Execution Trading¶

Practice notebooks for pandas data manipulation, tailored to a central execution trading role.

Notebooks¶

• pandas-intermediate-practice.ipynb — Intermediate: Core ops, GroupBy, time series, merging, order book intro
• pandas-advanced-practice.ipynb — Advanced: Multi-index, volume-time bars, microstructure, performance, capstone
• Notebook answers: intermediate and advanced

Reference¶

• Pandas cheatsheet — concise pandas reference

How to use¶

1. Open a notebook in VS Code or Jupyter.
2. Run the Setup cell first, then each Synthetic Data cell before its section.
3. Each exercise is a markdown prompt followed by an empty code cell — write your answer there.
4. The advanced notebook's capstone includes expected numeric ranges for sanity-checking (not full solutions).

Prerequisites¶

• Python 3.8+
• pandas ≥ 1.5, numpy ≥ 1.21
• Basic familiarity with DataFrames and Series (for intermediate); groupby/rolling/merge (for advanced)

Topics covered¶

• Core DataFrame/Series ops — indexing, filtering, dtypes, missing data, sorting, string ops
• GroupBy & aggregation — multi-key grouping, transform, filter, pivot/melt, crosstab
• Time series & rolling — DatetimeIndex, resample, rolling/expanding, shift, EWMA, tz handling
• Merging & joining — merge types, concat, join, merge_asof for trade/bar alignment
• Order book / microstructure — L1/L2 snapshots, spread, mid-price, VWAP/TWAP, slippage, OFI, realized spread
• Performance & optimization — vectorization, eval/query, categoricals, downcasting, memory profiling, chunked processing

All datasets are generated synthetically in-notebook — no external files required.