Validation: Save and restart behavior

Recorded evidence: 5 variations; 5 passed, 0 failed, and 0 skipped (100.0% pass rate over all recorded variations).

Evidence boundary

The current checked-in campaign uses small synthetic or already-present inputs and does not contact NEON. It validates software contracts and diagnostics, not real-flightline scientific accuracy.

What this module test exercises

Write one ENVI cube in separate row chunks, reconstruct it exactly, and prove that reading a reusable artifact does not modify it.

Implementation exercised: EnviWriter plus deterministic artifact hashing

Inputs varied

Field	Why it is recorded
shape_y_x_b	Varies spatial dimensions and band count.
chunk_split_row	Moves the internal write boundary.

Checks and how to interpret them

Check	Question	PASS means	If it does not pass
chunked_write_reconstructs_cube	Do independent writes reconstruct the original cube?	Maximum absolute read-back error is zero.	Inspect file offsets, axis order, and partial-write bounds.
expected_byte_count	Is the binary artifact complete for its declared shape and dtype?	File bytes equal the exact float32 shape-derived count.	Treat mismatch as truncation, padding, or header/data disagreement.
read_does_not_mutate_artifact	Is inspection itself non-destructive?	SHA-256 is identical before and after read-back.	Investigate accidental write mode or hidden repair behavior.

Diagnostics recorded for every variation

Field	Why it is recorded
chunk_split_row	Boundary between the two writes.
image_bytes	Observed binary size.
max_absolute_error	Largest reconstruction difference.
sha256	Post-read content digest.

Input variations and results

On narrow screens, scroll the table horizontally to see every diagnostic and check.

Variation	Input variation	Result	Diagnostics	Explicit checks
save_restart-001 Write an ENVI cube in two chunks split at row 1.	chunk_split_row=1; shape_y_x_b=[3,4,2]	PASS	chunk_split_row=1; image_bytes=96; max_absolute_error=0; sha256=aa7d1418648d4f23fc6002a9741b868f27493e546cc7903b635f03…	chunked_write_reconstructs_cube=✓; expected_byte_count=✓; read_does_not_mutate_artifact=✓
save_restart-002 Write an ENVI cube in two chunks split at row 2.	chunk_split_row=2; shape_y_x_b=[4,5,3]	PASS	chunk_split_row=2; image_bytes=240; max_absolute_error=0; sha256=c556b17b18f405119ce671fb392c8988c9dfc84c60104b2fdfe571…	chunked_write_reconstructs_cube=✓; expected_byte_count=✓; read_does_not_mutate_artifact=✓
save_restart-003 Write an ENVI cube in two chunks split at row 3.	chunk_split_row=3; shape_y_x_b=[5,6,4]	PASS	chunk_split_row=3; image_bytes=480; max_absolute_error=0; sha256=308faef52f0c097d825d98b121daf2993679072f717762316edba3…	chunked_write_reconstructs_cube=✓; expected_byte_count=✓; read_does_not_mutate_artifact=✓
save_restart-004 Write an ENVI cube in two chunks split at row 4.	chunk_split_row=4; shape_y_x_b=[6,4,5]	PASS	chunk_split_row=4; image_bytes=480; max_absolute_error=0; sha256=a2ef06292a890769c6d044bcbdb57ee490c98c31d66514ff1b05a6…	chunked_write_reconstructs_cube=✓; expected_byte_count=✓; read_does_not_mutate_artifact=✓
save_restart-005 Write an ENVI cube in two chunks split at row 1.	chunk_split_row=1; shape_y_x_b=[3,5,6]	PASS	chunk_split_row=1; image_bytes=360; max_absolute_error=0; sha256=05dad747989a20be5ab54817a75191c92e2e68ca470e6c3c9f41e2…	chunked_write_reconstructs_cube=✓; expected_byte_count=✓; read_does_not_mutate_artifact=✓

What a passing result establishes

Deterministic chunk writes, complete files, and non-mutating reads.

What it does not establish

Recovery from process termination during a write or distributed-filesystem failures.

The matching real stage checks are explained in the stage QA test guide.

Example from the real R10C test run

Restart-safe stage files allow the combined report to compare persisted outputs without rerunning corrections.

The figure is evidence from one completed flightline, not a replacement for the variation table above. Open the real flightline walkthrough for exact values and limitations.

Expansion to 100 real variations

The repository includes a live 100-flightline campaign specification. It requires a pinned inventory of real flightline IDs plus an explicit compute, storage, and network allocation. Live results must be stored as a new campaign record; they must not overwrite this offline baseline.

Reproduce or expand this module

# Fast local evidence matrix (five variations per module)
python scripts/run_validation_campaign.py --iterations-per-module 5

# Exercise 100 deterministic small-data variations per module
python scripts/run_validation_campaign.py --iterations-per-module 100 \
  --output validation/results/offline-contract-100.json

python scripts/generate_validation_docs.py

The 100-case offline command scales contract variation and randomized synthetic inputs. It does not substitute for 100 distinct NEON downloads.

Last updated: 2026-08-14