Follow the transformation
Read inputs from declared locations rather than manually edited in-memory data.
Place transformations in deterministic functions where practical.
Parameterise dates/paths instead of changing source code between runs.
Reporting & Reproducibility · Lesson 86
Reproducible Scripts
A reproducible analysis script can be rerun from declared inputs to regenerate the same transformations and outputs.
Reproducible Scripts
A reproducible analysis script can be rerun from declared inputs to regenerate the same transformations and outputs. It avoids hidden notebook state and manual spreadsheet edits by making paths, parameters, dependencies and execution order explicit.
Learning goal: explain why Reproducible Scripts behaves this way, apply it to a small example, and verify the result independently. Begin by being able to justify this first step: Read inputs from declared locations rather than manually edited in-memory data.
Deeper walkthrough
Read Reproducible Scripts as a mechanism, not a recipe
Treat this as a sequence of observable decisions rather than one opaque command. Stage 1: Read inputs from declared locations rather than manually edited in-memory data. Stage 2: Place transformations in deterministic functions where practical. Stage 3: Parameterise dates/paths instead of changing source code between runs. Final checkpoint: Record dependencies and rerun in a clean environment.
Know what would convince you
- Recompute one result from a handful of source rows or an independent formula.
- Check row counts, group totals and units before interpreting differences.
Useful distinctionDefinition: The exact metric/selection/comparison being computed.
Click a stage to inspect what happens, what changes, and what should be checked before moving on.
Stage 1
Read inputs from declared locations rather…
Read inputs from declared locations rather than manually edited in-memory data. This is an input-preparation stage for Reproducible Scripts. Verify the relevant type, shape, units, keys, missingness or assumptions before later steps depend on them.
Input focus: confirm the data/object, units, type, shape and assumptions before the next operation depends on them.
Trace the mechanism step by step
- Read inputs from declared locations rather than manually edited in-memory data.
- Place transformations in deterministic functions where practical.
- Parameterise dates/paths instead of changing source code between runs.
- Write outputs to known locations with versionable names.
- Record dependencies and rerun in a clean environment.
Reproducible entry point
# Step 1 — Define the reusable `main` function; its indented body describes what happens for each call.
def main(input_path, output_path):
# load -> validate -> analyse -> save
# Step 2 — Execute this statement and inspect how it changes the current value, object or program state.
...
# Step 3 — Evaluate this condition and execute the indented branch only when the condition is true.
if __name__ == "__main__":
# Step 4 — Leave this block intentionally empty as a placeholder for later behaviour.
passExpected / illustrative result
A single declared entry point makes the execution order explicit; real code would fill in each validated stage.
Interpret the result.
For Reproducible Scripts, identify exactly what each reported quantity represents, including its units/denominator, and independently recompute one part of the result.
Place the concept correctly
DefinitionThe exact metric/selection/comparison being computed.EvidenceTable, formula or visual that answers the question.AuditIndependent count/total/rule check that can reveal an error.When it is appropriate
Use Reproducible Scripts when it answers a defined question in Reporting & Reproducibility and its inputs/assumptions match the current data or program state.
When to stop or reconsider
Reconsider Reproducible Scripts when the required information is unavailable, the operation would violate a validation/data boundary, or a simpler operation answers the question more transparently.
Failure modes to recognise
- Changing the population/grain without noticing it.
- Using an undefined denominator, time window, unit or category rule.
- Presenting a number/plot without reconciling it to source counts or totals.
How to check the result
- Recompute one result from a handful of source rows or an independent formula.
- Check row counts, group totals and units before interpreting differences.
- Change one source value and predict which reported value/mark should change.
Demonstrate understanding
Try this:
Construct a tiny example of Reproducible Scripts. First read inputs from declared locations rather than manually edited in-memory data. Then place transformations in deterministic functions where practical. Predict the result before execution and explain one boundary or failure case.
Use a very small example and calculate one quantity manually. Separate sample evidence from population/causal claims.
Check reasoning, not memorisation
Which approach best demonstrates understanding of Reproducible Scripts?
Remember the logic
Step 1Read inputs from declared locations rather than manually edited in-memory data.Step 2Place transformations in deterministic functions where practical.Step 3Parameterise dates/paths instead of changing source code between runs.Step 4Write outputs to known locations with versionable names.What to remember
- A reproducible analysis script can be rerun from declared inputs to regenerate the same transformations and outputs. It avoids hidden notebook state and manual spreadsheet edits by making paths, parameters, dependencies and execution order explicit.
- Read inputs from declared locations rather than manually edited in-memory data.
- Changing the population/grain without noticing it.
- Recompute one result from a handful of source rows or an independent formula.