/Users/sean/workspace/ContribSLIMpy/slimpy_contrib/ana/problems/l1_minimization_problem.py

Go to the documentation of this file.

"""
Perform 2D de-noiseing using landweber method.

@note that, as of August 28, 2006:
 - dimension fixed (2D)

@par REQUIREMENTS:
- Madagascar with SLIM toolbox
- SLIMpy
    
@param    lambdaMax=LMAX    [Default=0.01]
@param    lambdaMin=LMIN    [Default=0.6]
@param    OuterN=NBOUT      Number of outer loops [Default=2]
@param    InnerN=NBIN       Number of inner loops [Default=2]
"""

__copyright__ = """
Copyright 2008 Sean Ross-Ross
"""

__license__ =  """
This file is part of SLIMpy .

SLIMpy is free software: you can redistribute it and/or modify
it under the terms of the GNU Lesser General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.

SLIMpy is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public License
along with SLIMpy . If not, see <http://www.gnu.org/licenses/>.
"""



from SLIMpy import vector
from SLIMpy.linear_operators import surf, Identity
from slimpy_base.utils.slim_decorators import depends_on
from slimpy_contrib.ana.GenLandweber import GenThreshLandweber
from slimpy_contrib.ana.utils.thresholds import LinearCooling


__acknowledgments__ = """
Author:         G. Hennenfent
                Seismic Laboratory for Imaging and Modeling (SLIM)
                Department of Earth & Ocean Sciences (EOS)
                The University of British Columbia at Vancouver (UBC)

Acknowledgments: 
                Sean Ross-Ross
                Seismic Laboratory for Imaging and Modeling (SLIM)
                Department of Earth & Ocean Sciences (EOS)
                The University of British Columbia at Vancouver (UBC)

                Darren Thomson
                Seismic Laboratory for Imaging and Modeling (SLIM)
                Department of Earth & Ocean Sciences (EOS)
                The University of British Columbia at Vancouver (UBC)

Funding:        This work was carried out as part of the SINBAD project with
                financial support, secured through ITF, from the following
                organizations: BG Group, BP, Chevron, ExxonMobil, and Shell. SINBAD is
                part of a collaborative research and development grant (CRD) number
                334810-05 funded by the Natural Science and Engineering Research
                Council (NSERC)

"""



Get = lambda env, item: env.get( item ) or env.get( 'problem', env.get('default_pack',{}) ).get(item)


@depends_on( 'lambdaMax', 'lambdaMin' )
@depends_on( 'OuterN', 'InnerN' )
@depends_on( 'x0' )
def solver_callback( target, source, env, A ):
    """
    solver_callback( target, source, env, space ) -> Solver
    Set the default value for the solver. Solver defaults 
    to the thresholded landweber with a linear cooling 
    threshold scheme.
    @return Thresholded Landweber Solver with LinearCooling scheme
    """
    
    space = A.domain()
    
    lambdaMax = Get( env,'lambdaMax' )
    lambdaMin = Get( env,'lambdaMin'  )
    outer = int( Get( env,'OuterN' ) )
    inner = int( Get( env,'InnerN'  ) )
    
    thresh = LinearCooling( float( lambdaMax ), float( lambdaMin ), outer )
    I = Identity( space )
    
    
    x0 = env.get( 'x0' , None )
    if x0 is None:
        x0 = space.zeros()
    else:
        x0 = vector( x0  )
    
    solver = GenThreshLandweber( outer, inner, thresh, I, x0 )
    
    return solver

@depends_on(  'transparams' )
def transform_callback( target, source, env,  space ):
    '''
    transform_callback( target, source, env,  space ) -> A
    set the default value for the transform
    @return the inverse surfacelet transform
    '''
    transparams = Get( env,'transparams' )    
    A = surf( space ).inv()
    
    return A


def precondition_callback(target,source,env, space ):
    """
    precondition_callback(target,source,env, space) -> Identity(space)
    Default preconditioner defaults to the Identity 
    """
    return Identity(space)


l1_min = {}
l1_min['name'] = 'l1 minimizeation with landweber method'
l1_min['doc'] = {}


l1_min['precondition_callback'] = precondition_callback
l1_min['transform_callback'] = transform_callback
l1_min['solver_callback'] = solver_callback
l1_min['lambdaMax'] = .01
l1_min['doc']['lambdaMax']  = """
        This parameter must be greater 
        than lambdaMin. The closer lambdaMax is to the max value of the coefficients of the vector the more your 
        acurate your result will be, but the problem needs to have 
        more iterations. less iterations need a smaller value, 
        however your results will be less acurate.    
"""
l1_min['lambdaMin'] = .60
l1_min['doc']['lambdaMin']  = """
        The closer the 
        parameter is to 0 the more noise and signal is kept in the 
        result as lambdaMin increases more noise will be removed. 
"""
l1_min['OuterN'] = 2
l1_min['doc']['OuterN']  = """
        number of outer iterations for the solver
"""
l1_min['InnerN'] = 2
l1_min['doc']['InnerN']  = """
        number of inner iterations for the solver
"""
l1_min['transparams'] = [4, 32, 1]
l1_min['doc']['transparams']  = """
        default parameters to pass to the curvelet transform
"""
#l1_min['lettype']     = 'curve'
#l1_min['doc']['lettype']  = """
#        default transform should be either 'surf' or 'curve'
#"""

---