Simulation parameters¶

All parameters below are keyword arguments to master_pattern() and master_pattern_from_cif(). Defaults match the API.

What drives the pattern¶

Beam and specimen – voltage_kv, sigma_deg, omega_deg, energy_binwidth_keV

Monte Carlo – mc_backend, mc_auto_stop, mc_relative_tol, n_trajectories

Dynamical diffraction – bethe_c_strong, bethe_c_weak, bethe_c_cutoff, dbdiff_sg_cutoff, rank, dmin, exact_slow_cpu, verbosity

Raster – halfw (Lambert half-width; side length 1 + 2*halfw)

Structure – lattice, sites, b_iso

Beam and Monte Carlo¶

Parameter	Default	Role
`voltage_kv` `sigma_deg` `omega_deg` `energy_binwidth_keV` `relative_image_stop` `marginal_coverage` `mc_backend` `mc_auto_stop` `mc_relative_tol` `n_trajectories`	`20.0` `70.0` `0.0` `1.0` `0.01` `1.0` surrogate `True` `0.01` 1048576	Nominal beam energy (kV) Specimen tilt (degrees) Azimuthal sample rotation (degrees) Energy-loss bin width (keV) Stop bins when delta-image/image is low Fraction of MC energy bins to integrate `"gpu"` for full boundary Monte Carlo GPU MC: stop when fits converge GPU MC convergence tolerance GPU MC budget when `mc_auto_stop=False`

Dynamical diffraction¶

The solve uses the Bloch formulation. Bethe perturbation theory ranks reflections (bethe_c_* cutoffs).

Parameter	Default	Role
`bethe_c_strong` `bethe_c_weak` `bethe_c_cutoff` `dbdiff_sg_cutoff` `rank` `dmin` `halfw` `exact_slow_cpu` `verbosity`	`20.0` `40.0` `200.0` `1.0` `20` `0.05` `250` `False` `0`	Strong-beam excitation-score threshold Weak-beam threshold band Upper cutoff; beams above are excluded Double-diffraction excitation admission Smith / Lyapunov truncation rank Minimum d-spacing (nm) Lambert half-width (501 by 501 at default) Full-rank CPU Lyapunov (`numpy.linalg.eig`) Progress: 0 silent, 1 bins, 2 chunks

Display scaling: lambert_data() with normalize="minmax" or normalize="robust".

Debye-Waller factor¶

When b_iso is omitted on an Atom, or when a CIF lacks _atom_site_B_iso_or_equiv / _atom_site_U_iso_or_equiv, ebsdsim defaults to 0.5 A^2 (stored as 0.005 nm^2). Override per site when better values are known:

es.Atom("Ga", x=1/3, y=2/3, z=0.0, b_iso=0.45)

Saved `.npz` layout¶

Array	Meaning
`fundamental_sector` `fundamental_kij` `fundamental_khat` `pg_operators` `fs_normals` `bin_voltages_kv` `bin_weights` `site_weights` `meta_json`	Raw FS intensities `(E, S, n_k)` Lambert indices per FS pixel Unit directions per FS pixel Point-group rotation matrices Fundamental-sector bounding normals Dynamical voltage per energy bin MC weight per energy bin Site marginal weights Simulation metadata (UTF-8 JSON)

Offline loading uses ebsdsim.mploader (NumPy only). See Quick start and examples/02_save_and_load.ipynb.