🧪 Lab Environment, Data & Setup
Drive folder: https://drive.google.com/drive/folders/1QyGrbwbPByf79bZNmJ_dtmftOpObLym3
Environment Blueprint: Google Doc · Drive PDF
Synthetic Data & Defect Blueprint: Google Doc · Drive PDF
Online environment pages
Live GitHub lab workspace
The editable, rerunnable lab assets live outside the MkDocs docs/ tree in the private repository's /lab workspace on GitHub.
SQL setup
00_meridian_lab_bootstrap.sql— creates the pedagogical raw/staging/core/master/reference/warehouse/metadata/quality/audit schemas.01_environment_roles_template.sql— documents the planned training roles without hiding GRANT/REVOKE learning inside automation.
Python
meridian_data_generator.py— deterministic synthetic-data generator using seed42001.
Source-data control
lab/data/raw/README.md— canonical baseline rules and row counts.defect_map.csv— stable intentional defect IDs and learning uses.validation_summary.json— expected baseline counts and validation results.source_manifest.csv— file sizes, SHA-256 hashes, seed, and generation timestamp.
Frozen Drive archive
The original generated baseline ZIP remains in Drive as a frozen archival copy: Meridian_Synthetic_Source_Data_Baseline.zip.
Drive working areas
| Area | Drive folder | Purpose |
|---|---|---|
| 01 — Setup & Installation | Open | Local environment/setup artifacts. |
| 02 — Synthetic Data | Open | Canonical raw synthetic sources and manifests. |
| 03 — SQL Scripts | Open | Bootstrap, schema, validation and chapter SQL. |
| 04 — Python Scripts & Notebooks | Open | Data generation/profiling/validation/automation. |
| 05 — Diagrams | Open | Architecture, lineage, models and other visuals. |
| 06 — Generated Outputs & Evidence | Open | Results produced by labs. |
| 07 — Rebuild, Backup & Recovery | Open | Reproducibility and recovery assets. |
Core rule
The lab environment must be safe, synthetic, reproducible and disposable. Preserve canonical raw data. Changes happen in database/staging/processed/evidence layers or through controlled defect-injection/rebuild mechanisms—not by quietly hand-editing the source until no one knows the starting state.
Repository boundary
Text assets that benefit from version control belong in GitHub. The current program source PDFs are now archived locally in the private repository; Google-native originals and selected frozen/large binary archives remain linked in Drive where appropriate.