API map for analysis plugin contributors
This reference page maps common contributor needs to the stable import surface and the nearest detailed documentation. Use it as a lookup bridge when writing or reviewing analysis plugins.
Warning
polyzymd.analyses._framework is private implementation infrastructure, not a
contributor import surface. Leading-underscore helper modules inside built-in
plugin packages are examples of package organization only; they are not public
APIs for contributor plugins.
Public API map
Use the public facades below as your default imports. They keep contributor plugins insulated from internal framework changes while still covering the common plugin lifecycle tasks.
Create a plugin class
Import lifecycle classes and comparison vocabulary from
polyzymd.analyses.base.
from polyzymd.analyses.base import Analysis, MetricValue
Use this surface when you need to:
subclass
Analysis;receive lifecycle contexts such as
AggregateContext,ComparisonContext, orPlotContext;describe scalar metrics with
MetricValue;return framework comparison models such as
ComparisonResult.
Build MDAnalysis-native replicate jobs
Import trajectory-native job helpers from polyzymd.analyses.mda.
from polyzymd.analyses.mda import FrameSelection, MDAAnalysisJob
Use this surface when you need to:
build replicate jobs around MDAnalysis
AnalysisBase-compatible work;adapt simple trajectory functions into framework jobs;
select frames with
FrameSelection;receive MDAnalysis job context objects from the framework.
Emit artifacts and sidecars
Import artifact, collector, and sidecar helpers from polyzymd.analyses.mda.
from polyzymd.analyses.mda import ArtifactStore, MDAArtifactCollector
from polyzymd.analyses.mda import MDACollectorContext, ReplicateArtifact
Use this surface when you need to:
convert completed MDAnalysis jobs into canonical replicate artifacts;
write and read artifact JSON and manifests;
attach validated sidecars with
ArtifactSidecarRef;use default replicate-to-condition aggregation;
compare condition artifacts without inventing plugin-specific cache files.
Discover and test plugins
Import plugin discovery helpers from polyzymd.analyses.
from polyzymd.analyses import get_analysis, list_analyses
Use this surface in tests, CLI helpers, and developer diagnostics when you need to confirm that a plugin is auto-discovered or retrieve a plugin class by name.
Compare scalar metrics
Import reusable scalar comparison and formatting helpers from
polyzymd.analyses.stats.
from polyzymd.analyses.stats import default_scalar_comparison
from polyzymd.analyses.stats import format_scalar_comparison
Use this surface when a plugin reports one or more scalar MetricValue entries
and can use the framework’s default t-tests, ANOVA, ranking, and CLI summary
formatting.
Start from scaffolding and configuration reference
Use the scaffold command when you want a working plugin skeleton and matching tests before filling in analysis-specific logic.
pixi run -e build polyzymd new-analysis solvent_shell
Replace solvent_shell with the desired plugin name.
See CLI new-analysis reference for command options.
See Analysis Plugin Settings Reference for built-in
comparison.yamlplugin settings underplugins.<plugin_name>.
Stable import boundaries
Prefer
polyzymd.analyses.basefor plugin lifecycle classes and comparison result vocabulary.Prefer
polyzymd.analyses.mdafor trajectory-native jobs, collectors, artifacts, sidecars, stores, aggregation, and comparison.Use only utilities documented on Analysis Shared Utilities from
polyzymd.analyses.shared.Treat built-in plugin package roots, such as
polyzymd.analyses.contacts, as public locations for their documented analysis classes and settings. Treat their leading-underscore helper modules as implementation examples, not contributor dependencies.