Composed Analyses
This section documents composed analyses in ReaxKit — higher‑level routines that combine data from multiple ReaxFF files to compute physically meaningful, structure‑aware, or system‑level quantities.
These analyses sit above per‑file analyses and represent the main scientific value layer of ReaxKit.
What “composed” means in ReaxKit
A composed analysis:
- Uses two or more ReaxFF files
- Combines data from multiple handlers
- Encodes domain logic (physics / chemistry / materials insight)
- Often operates on local environments, clusters, or time‑evolving structures
- Is exposed via CLI workflows that coordinate several handlers internally
Typical examples:
- Coordinates (xmolout) + connectivity (fort.7)
- Energies (energylog) + volume (fort.74)
- Trajectory (xmolout) + charges + connectivity → dipoles / polarization
Common composed analysis patterns
Structure + connectivity
- Local coordination environments
- Molecule / cluster identification
- Bond‑aware geometric properties
- Per‑atom or per‑cluster observables
Typical inputs
- xmolout
- fort.7
Trajectory + derived physics
- Dipole moments and polarization
- Mean‑square displacement by species or cluster
- Time‑resolved local observables
Typical inputs
- xmolout
- fort.7
- (optionally) charge or electrostatics data
Thermodynamics + mechanics
- Energy–volume relations
- Bulk modulus fits
- Stress–strain analysis
Typical inputs
- energylog / fort.73
- fort.74
- fort.76
- fort.99
What each composed analysis page contains
Each page typically documents:
- Scientific objective
- Required input files
- Handlers and analyzers used
- Data‑flow diagram (conceptual)
- Python API example
- CLI workflow usage
- Exported outputs
- Physical assumptions and caveats
When to start from per‑file analyses instead
If you only need:
- Raw trajectories
- Energies vs iteration
- Single‑file summaries or plots
→ Start in per-file analysis instead.
Composed analyses intentionally assume familiarity with ReaxFF outputs and build on clean, validated per‑file data.