Problem 1
Reconstruct the method from Lectures 03-04. Define input, transformation, output, and one validity condition. Use shared ASTR430 values and show units or decision criteria.
Objectives
- Build an auditable workflow.
- Use quantitative uncertainty or model-selection reasoning.
- Cite archive and software sources precisely.
- Communicate limitations and revision decisions.
Instructions
Submit four answers, code or reproducible pseudocode, one labeled figure or table, and a provenance note. Report units, assumptions, exact source anchors, and instructor-provided data. Revisions are encouraged with a change log.
Problem 2
Use data/gaia_query_manifest.csv to produce an audit table or figure. Include exact records, a reproducible calculation, and this sensitivity test: vary match radius from 0.5 to 1.0 arcsec.
Problem 3
Write a reviewer response to a claim that a shared notebook proves reproducibility. Identify four missing provenance fields and explain how each could change the conclusion.
Problem 4
Transfer: design a follow-up analysis or observing step that discriminates between two explanations. State source, measurable quantity, uncertainty budget, and failure condition. Provenance requirement: Record endpoint, ADQL, columns, release, constraints, row count, timestamp, and citation.
Deliverables
100 points: method 25, quantitative/data audit 30, critique 20, transfer design 25.