From 7d22974ca26f11815fbdb98d6ad122725dc8d08e Mon Sep 17 00:00:00 2001 From: Jammy2211 Date: Tue, 14 Jul 2026 08:48:25 +0100 Subject: [PATCH] docs: fix dead iterations_per_update -> iterations_per_quick/full_update MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit af.Nautilus (and the search base) has no iterations_per_update param — the real knobs are iterations_per_quick_update / iterations_per_full_update (0 vs 40/58 occurrences in PyAutoFit). The wiki's code example (searches.md) and prose (non_linear_search.md, analysis_objects.md) used the dead name, which silently no-ops if copied. Corrected all three; noted there is no iterations_per_update. Co-Authored-By: Claude Opus 4.8 Claude-Session: https://claude.ai/code/session_01HD3ZkpFRydshFLeu2V6xDi --- wiki/core/api/analysis_objects.md | 3 ++- wiki/core/api/searches.md | 2 +- wiki/core/concepts/non_linear_search.md | 5 +++-- 3 files changed, 6 insertions(+), 4 deletions(-) diff --git a/wiki/core/api/analysis_objects.md b/wiki/core/api/analysis_objects.md index fdf3131..f5d61d1 100644 --- a/wiki/core/api/analysis_objects.md +++ b/wiki/core/api/analysis_objects.md @@ -170,7 +170,8 @@ concise API reference. This is the ground the ## Visualisations during fit Analyses also produce diagnostic plots that PyAutoFit writes to `output/.../image/` -at each update step (controlled by `iterations_per_update` on the search). The +at each update step (controlled by `iterations_per_quick_update` / +`iterations_per_full_update` on the search). The plot routines live in the same packages as the analyses: - `PyAutoLens:autolens/imaging/model/visualizer.py` diff --git a/wiki/core/api/searches.md b/wiki/core/api/searches.md index b11b61e..b8d53c0 100644 --- a/wiki/core/api/searches.md +++ b/wiki/core/api/searches.md @@ -31,7 +31,7 @@ af.Nautilus( path_prefix=..., name=..., unique_tag=..., n_live=200, # 100 (exploration) / 200 (production) / 400+ (complex) number_of_cores=4, - iterations_per_update=2500, + iterations_per_quick_update=2500, # cadence for the on-the-fly max-likelihood fit ) ``` diff --git a/wiki/core/concepts/non_linear_search.md b/wiki/core/concepts/non_linear_search.md index c40f146..4552bf8 100644 --- a/wiki/core/concepts/non_linear_search.md +++ b/wiki/core/concepts/non_linear_search.md @@ -92,8 +92,9 @@ and of the SLaM pipeline. to produce a unique output directory. - **`unique_tag`** — extra identifier to distinguish otherwise-identical runs. - **`number_of_cores`** — parallel likelihood evaluations. -- **`iterations_per_update`** — how often the search writes intermediate samples to - disk. +- **`iterations_per_quick_update`** / **`iterations_per_full_update`** — how often the + search writes the on-the-fly max-likelihood fit (quick) vs the full output — samples, + visuals, `model.result` (full). There is no `iterations_per_update`. Per-search knobs (`n_live` for nested, `nwalkers` / `nsteps` for MCMC) are in the search-specific entries of [`../api/searches`](../api/searches.md).