MiRAParam < Jmmc/Software

---+++ Data selection:
   *   -target=NAME                            Name of the astrophysical object
   *   -effwave=LENGTH                         Effective wavelength, e.g. 1.6micron
   *   -effband=LENGTH                         Effective bandwidth, e.g. 200nm
   *   -wavemin=LENGTH                         Minimum wavelength, e.g. 1.5µm
   *   -wavemax=LENGTH                         Maximum wavelength, e.g. 1.7µm


---+++ Which data to fit and how:
   *   -use_vis=all|none|amp|phi               Complex visibility data to consider
   *   -use_vis2=all|none                      Powerspectrum data to consider
   *   -use_t3=all|none|amp|phi                Bispectrum data to consider
   *   -convexcost=yes|no                      Use convex approximation for fitting complex data?
   *   -phasecost=vonmises|haniff|convexlimit  Co-log-likelihood approximation to use for phase data


---+++ Image parameters and direct model:
   *   -pixelsize=ANGLE                        Angular size of pixels, e.g. 0.1mas
   *   -fov=ANGLE                              Angular size of the field of view, e.g. 20mas
   *   -imagesize=DIM|DIM1xDIM2                Number of pixels per side of the image
   *   -flux=VALUE                             Total flux (sum of pixels = VALUE)
   *   -fluxerr=VALUE                          Assumed standard deviation for the total flux
   *   -min=LOWER                              Lower bound for the pixel values
   *   -max=UPPER                              Upper bound for the pixel values
   *   -xform=nfft|separable|nonseparable      Method to compute the Fourier transform
   *   -smearingfunction=none|sinc|gauss       Method to model the effects of the spectral bandwidth smearing
   *   -smearingfactor=VALUE                   Factor to scale the effects of the spectral bandwidth smearing
   *   -nthreads=NUMBER                        Number of threads for the fast Fourier transform


---+++ Regularization settings:
   *   -regul=NAME                             Name of regularization method (-regul=help for more information)
   *   -mu=VALUE(S)                            Global regularization weight
---++++  Available regularizations:
   *   -regul=hyperbolic -tau=...  Edge-preserving smoothness with hyperbolic norm, approximates total variation (TV) with TAU very small.
      *   -tau=VALUE                              Edge preserving threshold
      *   -eta=VALUE(S)                           Gradient scales along dimensions
   *   -regul=compactness -gamma=...  Quadratic compactness with full-width at half maximum GAMMA.
      *   -gamma=FWHM                             A priori full half width at half maximum, e.g. 15mas


---+++ Initial image:
   *   -initial=NAME                           FITS file or method for initial image
   *   -seed=VALUE                             Seed for the random generator


---+++ Output file:
   *   -save_visibilities                      Save model complex visibilities
   *   -overwrite                              Overwrite output if it exists
   *   -bitpix=BITPIX                          Bits per pixel
   *   -save_initial                           Save initial image as a secondary HDU in the result


---+++ Reconstruction strategy:
   *   -bootstrap=COUNT                        Number of bootstrapping iterations
   *   -recenter                               Recenter result of bootstrapping iterations
   *   -threshold=FRACTION                     Level for soft-thresholding input image(s)


---+++ Messages:
   *   -quiet                                  Suppress most messages
   *   -verb=COUNT                             Verbose level


---+++ Optimizer Settings:
   *   -mem=COUNT                              Number of previous steps to memorize in VMLMB
   *   -ftol=REAL                              Function tolerance for the global convergence
   *   -gtol=REAL                              Gradient tolerance for the global convergence
   *   -maxiter=COUNT                          Maximum number of iterations
   *   -maxeval=COUNT                          Maximum number of evaluations of the objective function


---+++     Line search:
   *   -sftol=REAL                             Function tolerance for the line search
   *   -sgtol=REAL                             Gradient tolerance for the line search
   *   -sxtol=REAL                             Step tolerance for the line search


---+++ Miscellaneous:
   *   -settings=NAME                          Name of the FITS parameter file
   *   -plugin=NAME                            Name of plugin
   *   -debug                                  Debug mode
   *   -help                                   Print out this help
   *   -version                                Print out version number
This topic: Jmmc/Software > WebHome > MFIR > MfirSprint1 > OImagingParam > MiRAParam
Topic revision: r1 - 2018-09-21 - FerreolSoulez