PyPI page
Author:
Hemant Sharma
License:
BSD-3-Clause
Summary:
Joint differentiable powder + FF-HEDM detector calibration for MIDAS. Breaks single-distance per-panel rank-1 degeneracy by combining a powder calibrant pseudo-strain residual with HEDM grain spot residuals in one Bayesian MAP + Laplace inference. Pure-Python, fully autograd-traced; consumes the shared midas_peakfit substrate (ParameterSpec, lm_minimise, laplace_at_map, TPSpline, zero_sum_residual) plus midas_calibrate_v2 (powder forward + pseudo_strain) and midas_fit_grain (HEDM spot forward).
Latest version:
0.6.0
Required dependencies:
midas-calibrate
|
midas-calibrate-v2
|
midas-diffract
|
midas-fit-grain
|
midas-hkls
|
midas-params
|
midas-peakfit
|
midas-process-grains
|
numpy
|
scipy
|
torch
Optional dependencies:
pytest
Downloads last day:
13
Downloads last week:
178
Downloads last month:
474
7-day smoothing