Skip to content

FMC alignment figure: one legend, and the multiblock PNGs refreshed to 1.83.0 - #93

Merged
kgdunn merged 1 commit into
mainfrom
claude/issues-154-156-process-3cu4iq
Sep 8, 2026
Merged

kgdunn merged 1 commit into
mainfrom
claude/issues-154-156-process-3cu4iq

Conversation

@kgdunn

@kgdunn kgdunn commented Sep 8, 2026

Copy link
Copy Markdown
Owner

Follows #92. Two changes to the FMC batch case-study figures.

The alignment figure

The legend sat in the upper left of the "before" panel, over the rising temperature curves. It now appears once, in the lower right of the warp panel, where the clock-time lines run from the lower left and leave that corner empty. The left and middle panels carry no legend.

The three highlighted batches keep their colours across all three panels, so one legend serves all of them. Its labels now carry each batch's duration (batch 9: 93 units, batch 2: 128 units, batch 34: 200 units), which reads directly off that panel's y-axis and corroborates the caption's "93 and 200 time units".

Four multiblock PNGs refreshed

block-scores, batch-mbpls, mbpls-z and anomalous also change. They were last generated before the pooled super-score fix in process-improve 1.83.0, while the chapter's numbers were updated for 1.83.0 in pid-book#273. The figures and the prose had been out of step; regenerating brings them back into line.

The changes are small and the readings hold:

  • the chemistry block still keeps 6.5% on the first component, the number the chapter quotes;
  • the four batches classed good still sit with the abnormal group in the trajectory block and with the good group in both initial-condition blocks, which is the point of block-scores.

Generated against process-improve at origin/main, whose src/ is byte-identical to the released 1.83.0, so these match what a reader gets from PyPI. The script docstring now declares 1.83.0 and says why the multiblock figures need it.

Verification

  • Regenerated the full FMC figure set; only the five PNGs above differ.
  • Viewed the alignment figure: the legend clears every curve, and no label collides.
  • No other case-study script touched.

Merge before pid-book#273, as the book's PDF step resolves images from this repository's default branch.

🤖 Generated with Claude Code

https://claude.ai/code/session_01LXthGpHLQFfGubBiKFtGAE


Generated by Claude Code

The legend sat in the upper left of the "before" panel, over the rising
temperature curves. It now appears once, in the lower right of the warp
panel, where the clock-time lines run from the lower left and leave that
corner empty. The three batches keep their colours across the panels, so
one legend serves all three, and its labels carry each batch's duration,
which reads directly off that panel's y-axis.

The four multiblock PNGs also change. They were last generated before the
pooled super-score fix and were stale against the chapter, whose numbers
were updated for process-improve 1.83.0. Regenerating brings them into
step: the chemistry block still keeps 6.5% on the first component, as the
chapter says, and the four batches classed good still sit with the abnormal
group in the trajectory block and with the good group in both
initial-condition blocks. The script now declares 1.83.0 and says why.

Generated against process-improve at origin/main, whose src/ is identical
to the released 1.83.0.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01LXthGpHLQFfGubBiKFtGAE
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants