Financial Computing Python
objective
Build Python tooling for quant workflows with reproducible data transformations and testable research code.
workflow
- •define data contracts, calendar conventions, and feature definitions before coding.
- •implement pandas pipelines for ingestion, cleaning, and time alignment.
- •build reusable modules for signals, portfolio logic, and analytics.
- •validate outputs with deterministic unit tests and replay datasets.
- •promote only after profiling runtime and memory behavior on production-size data.
required diagnostics
- •feature leakage and look-ahead checks on timestamp alignment.
- •missing-data handling consistency across symbols and sessions.
- •run-time and memory cost by pipeline stage.
- •signal reproducibility across repeated runs and seeds.
- •backtest consistency between batch and incremental updates.
risk controls
- •enforce schema validation at ingestion boundaries.
- •enforce test coverage for pricing, returns, and pnl calculations.
- •enforce rollback to previous release on reproducibility failures.
outputs
- •run
python scripts/financial_computing_python_diagnostics.py input.csv --output diagnostics.jsonand keep the json artifact. - •write an implementation memo using
references/financial-computing-python-playbook.mdwith assumptions, tests, limits, and rollout plan.
resources
- •use
scripts/financial_computing_python_diagnostics.pyfor deterministic diagnostics. - •use
references/financial-computing-python-playbook.mdfor the domain checklist and delivery structure.