pymerlin.reg.setup_optimizer

pymerlin.reg.setup_optimizer(PixelType, opt_range, relax_factor, nit=250, learning_rate=0.1, convergence_window_size=10, convergence_value=1e-06, min_step_length=0.0001)[source]

Setup optimizer object

Parameters

  • PixelType (itkCType) – ITK pixel type

  • opt_range (list) – Range for optimizer

  • relax_factor (float) – Relaxation factor

  • nit (int, optional) – Number of iterations. Defaults to 250.

  • learning_rate (float, optional) – Optimizer learning rate. Defaults to 0.1.

  • convergence_window_size (int, optional) – Number of points to use in evaluating convergence. Defaults to 10.

  • convergence_value ([type], optional) – Value at which convergence is reached. Defaults to 1E-6.

Returns

Optimizer object

Return type

itk.RegularStepGradientDescentOptimizerv4