Practice Map¶
Reference Position¶
flowchart TD
family["Reproducible Research"] --> program["Deep Dive DVC"]
program --> reference["Practice Map"]
reference --> review["Design or review decision"]
review --> capstone["Capstone proof surface"]
flowchart TD
trigger["Hit a naming, boundary, or trade-off question"] --> lookup["Use this page as a glossary, map, rubric, or atlas"]
lookup --> compare["Compare the current code or workflow against the boundary"]
compare --> decision["Turn the comparison into a keep, change, or reject call"]
Read the first diagram as a lookup map: this page is part of the review shelf, not a first-read narrative. Read the second diagram as the reference rhythm: arrive with a concrete ambiguity, compare the current work against the boundary on the page, then turn that comparison into a decision.
The course should make it obvious what to inspect, what to run, and what success looks like at each stage.
This page collects that information in one place.
Module Practice Surfaces¶
| Module | Primary practice surface | Main proof loop | Best capstone follow-up |
|---|---|---|---|
| 01 | thought experiments and failure analysis | identify hidden inputs and missing evidence | inspect README.md and TOUR.md as a contract specimen |
| 02 | state-layer inspection | compare workspace, Git, lock, cache, and remote roles | inspect dvc.lock and promoted outputs |
| 03 | environment declaration choices | compare lockfiles, containers, and CI authority | inspect pyproject.toml and repository execution targets |
| 04 | pipeline declaration review | inspect dvc.yaml, rerun behavior, and dvc.lock |
inspect stage edges in the capstone |
| 05 | params and metrics review | compare declared controls and tracked outcomes | inspect params.yaml, tracked metrics, and publish evidence |
| 06 | experiment discipline | vary params without corrupting baseline state | run or inspect experiment-oriented surfaces |
| 07 | collaboration and CI expectations | compare human process to verification commands | inspect confirm, push, and recovery-drill |
| 08 | recovery and retention thinking | rehearse loss and restoration logic | inspect remote-backed recovery and promoted bundles |
| 09 | promotion boundary review | inspect publish contract and evidence bundle | inspect publish/v1/ and verify it |
| 10 | repository review | write a short stewardship review | use the capstone as the review specimen |
Three Reusable Proof Loops¶
State truth loop¶
Use when you want to inspect whether the repository is saying what changed honestly.
Comparison loop¶
Use when you want to inspect whether params and metrics remain meaningful across runs.
Recovery loop¶
Use when you want to inspect whether durability claims survive local loss.
make PROGRAM=reproducible-research/deep-dive-dvc capstone-recovery-drill
make PROGRAM=reproducible-research/deep-dive-dvc capstone-recovery-review
Best Study Habit¶
For each module:
- name the state boundary the module is teaching
- identify the proof command before you run it
- inspect the capstone only after the smaller concept is legible
- write down what would break if the contract were false
That keeps the course centered on comprehension instead of command accumulation.