Input

Parsing uses PQAnalysis. Formats are detected automatically; specify a flag when detection is ambiguous.

Output	Files	Explicit selection	Independent axis
PQ energy	.en and matching .info	--pq	Simulation time, with its recorded unit
QMCFC energy	.en and matching .info	--qmcfc	Simulation time or step
PQ cell	.box; rows step x y z alpha beta gamma	--box	Simulation step
PQ optimizer	.opt	--opt	Optimization step

pqenalyzer web --pq simulation.en
pqenalyzer web --box simulation.box
pqenalyzer web --opt optimization.opt

Energy metadata supplies labels and units. Single-column PQ metadata rows are supported. Box parameters include lengths, angles, and BOX-VOLUME. Optimizer parameters include energy changes, forces, convergence states, and limits. State -1 means not converged, 0 disabled, and 1 converged; the first optimizer row is the initialization snapshot. Check the PQ log for final completion status.

Dataset conventions

• Files are read in command-line order. A selected parameter uses only files containing that parameter, in that order.

• Shared parameter labels must have identical units. A mismatch is rejected; units are not converted.

• Finite, strictly increasing time across the complete sequence is used as the chart axis. A restart, overlap, or invalid time switches the axis to samples 1, 2, …, N; observations retain their order.

• Continuation segments should represent the same sampling process. Autocorrelation pairs cross file boundaries. Analyze independent replicas or changed ensembles in separate sessions when interpreting correlation and uncertainty.

Charts may reduce the displayed point count while preserving endpoints, bucket extrema, and representative missing-value gaps. Statistics and histogram counts use the full finite data. Missing values appear as gaps; the different analysis rules are listed in Analysis.

For reproducible results, retain input files and metadata, file order, software versions, sampling interval, selected parameter, and any chosen window or histogram settings.