1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
|
\section{Driver Classes}\label{chapter:ref:Drivers:Drivers}
The inversion minimizes an appropriate cost function $J$ to find the physical parameter distribution
(or more precisely the level set function) which gives the best fit to measured data. A
particular inversion case (gravity, magnetic or joint) is managed through
an instance of a specialization of the \class{InversionDriver} class. The task of the class instance
is to set up the appropriate cost function, to manage solution parameters and to run the optimization process.
\subsection{Template}
\begin{classdesc*}{InversionDriver}
template for inversion drivers.
\end{classdesc*}
\begin{methoddesc}[InversionDriver]{getCostFunction}{}
returns the cost function of the inversion. This will be an instance of the \class{InversionCostFunction} class, see Section~\ref{chapter:ref:inversion cost function}.
Use this method to access or alter attributes or call methods of the underlying cost function.
\end{methoddesc}
\begin{methoddesc}[InversionDriver]{getSolver}{}
returns the instance of the solver class used to minimize the cost function, see Chapter~\ref{chapter:ref:Minimization}.
Use this method to modify solver options.
\end{methoddesc}
\begin{methoddesc}[InversionDriver]{getDomain}{}
returns the domain of the inversion as an \escript \class{Domain} object.
\end{methoddesc}
\begin{methoddesc}[InversionDriver]{setSolverMaxIterations}{\optional{maxiter=\None}}
sets the maximum number of iteration steps for the solver used to minimize the cost function. The default value is 200.
If the maximum number is reached, the iteration will be terminated and \exception{MinimizerMaxIterReached} is thrown.
\end{methoddesc}
\begin{methoddesc}[InversionDriver]{setSolverTolerance}{\optional{m_tol=\None} \optional{, J_tol=\None}}
set the tolerance for the solver used to minimize the cost function. If \member{m_tol} is set the iteration is terminated
if the relative change of the level set function is less than or equal \member{m_tol}, see condition~\ref{EQU:MIN:3a}.
If \member{J_tol} is set the iteration is terminated if the change of the cost
function relative to the initial value is less than or equal \member{J_tol}, , see condition~\ref{EQU:MIN:3b}.
If both tolerances are set both stopping criteria need to be met.
By default \member{tol}=1e-4 and \member{J_tol}=\None.
\end{methoddesc}
\begin{methoddesc}[InversionDriver]{getLevelSetFunction}{}
returns the level set function as solution of the optimization problem.
This method can only be called if the optimization process as been completed.
If the iteration failed the last available approximation of the solution is returned.
\end{methoddesc}
\begin{methoddesc}[InversionDriver]{run}{}
this method runs the optimization solver and returns the physical parameter(s)
from the output of the inversion. Notice that the \method{setup} method must be
called before the first call of \method{run}.
The call can fail as the maximum number of iterations is reached in which case
a \exception{MinimizerMaxIterReached} exception is thrown or as there is an
incurable break down in the iteration in which case a \exception{MinimizerIterationIncurableBreakDown} exception is thrown.
\end{methoddesc}
\subsection{Gravity Inversion Driver}
For examples of usage please see Chapter~\ref{Chp:cook:gravity inversion}.
\begin{classdesc}{GravityInversion}{\optional{solverclass=None}
\optional{, fixGravityPotentialAtBottom=False}
}
Driver class to perform an inversion of Gravity (Bouguer) anomaly data.
This class is a sub-class of \class{InversionDriver}. The class uses the
standard \class{Regularization} for a single level set function,
see Chapter~\ref{Chp:ref:regularization}, \class{DensityMapping} mapping,
see Section~\ref{Chp:ref:mapping density}, and the gravity forward model
\class{GravityModel}, see Section~\ref{sec:forward gravity}.
\member{solverclass} set the solver class to be used for inversion,
see Chapter~\ref{chapter:ref:Minimization}.
By default the limited-memory Broyden-Fletcher-Goldfarb-Shanno (\emph{L-BFGS})~\cite{Nocedal1980}\index{L-BFGS} solver is used.
If \member{fixGravityPotentialAtBottom} is set \optional{, fixGravityPotentialAtBottom=False}to \True the gravity potential at the bottom is set to zero.
\end{classdesc}
\begin{methoddesc}[GravityInversion]{fixGravityPotentialAtBottom}{\optional{status=\True}}
If \member{status} is \True the gravity potential at the bottom is set to zero. Otherwise the gravity
potential at the top is set to zero only.
\end{methoddesc}
\begin{methoddesc}[GravityInversion]{setup}{
domainbuilder
\optional{, rho0=\None}
\optional{, drho=\None}
\optional{, z0=\None}
\optional{, beta=\None}
\optional{, w0=\None}
\optional{, w1=\None}
\optional{, rho_at_depth=\None}
}
sets up the inversion from an instance \member{domainbuilder} of a \class{DomainBuilder}, see Section~\ref{Chp:ref:domain builder}.
Only gravitational data attached to the \member{domainbuilder} are considered in the inversion.
\member{rho0} defines a reference density anomaly (default is 0),
\member{drho} defines a density anomaly (default is $2750 \frac{kg}{m^3}$),
\member{z0} defines the depth weighting reference depth (default is \None), and
\member{beta} defines the depth weighting exponent (default is \None),
see \class{DensityMapping} in Section~\ref{Chp:ref:mapping density}.
\member{w0} and \member{w1} define the weighting factors
$\omega^{(0)}$ and
$\omega^{(1)}$, respectively (see Equation~\ref{EQU:REG:1}).
By default \member{w0}=\None and \member{w1}=1 are used.
\member{rho_at_depth} sets the value for density at depth. This is only used if density is fixed below a certain depth,
see \class{Domain Builder} in Section~\ref{Chp:ref:domain builder}.
\end{methoddesc}
\begin{methoddesc}[GravityInversion]{setInitialGuess}{\optional{rho=\None}}
sets an initial guess for the density anomaly. By default zero is used.
\end{methoddesc}
\subsection{Magnetic Inversion Driver}
For examples of usage please see Chapter~\ref{Chp:cook:magnetic inversion}.
\begin{classdesc}{MagneticInversion}{\optional{solverclass=None}, \optional{self_demagnetization=\False}
}
Driver class to perform an inversion of magnetic anomaly data. This class
is a sub-class of \class{InversionDriver}. The class uses the standard
\class{Regularization} class for a single level set function, see Chapter~\ref{Chp:ref:regularization},
\class{SusceptibilityMapping} mapping, see Section~\ref{Chp:ref:mapping susceptibility},
and the magnetic forward model \class{SelfDemagnetizationModel} (if \member{self_demagnetization}=\True,
see Section~\ref{sec:forward self demagnetic}) or \class{MagneticModel}
(if \member{self_demagnetization}=\False, see Section~\ref{sec:forward magnetic}).
\member{solverclass} set the solver class to be used for inversion,
see Chapter~\ref{chapter:ref:Minimization}.
By default the limited-memory Broyden-Fletcher-Goldfarb-Shanno (\emph{L-BFGS})~\cite{Nocedal1980}\index{L-BFGS} solver is used.
\end{classdesc}
\begin{methoddesc}[MagneticInversion]{fixMagneticPotentialAtBottom}{\optional{status=\True}}
If \member{status} is \True the magnetic potential at the bottom is set to zero. Otherwise the magnetic
potential at the top is set to zero only.
\end{methoddesc}
\begin{methoddesc}[MagneticInversion]{setup}{
domainbuilder
\optional{, k0=\None}
\optional{, dk=\None}
\optional{, z0=\None}
\optional{, beta=\None}
\optional{, w0=\None}
\optional{, w1=\None}
\optional{, k_at_depth=\None}
}
sets up the inversion from an instance \member{domainbuilder} of a \class{DomainBuilder}, see Section~\ref{Chp:ref:domain builder}.
Only magnetic data attached to the \member{domainbuilder} are considered in the inversion.
\member{k0} defines a reference susceptibility anomaly (default is $0$),
\member{dk} defines a susceptibility anomaly scale (default is $1$),
\member{z0} defines the depth weighting reference depth (default is \None), and
\member{beta} defines the depth weighting exponent (default is \None),
see \class{SusceptibilityMapping} in Section~\ref{Chp:ref:mapping susceptibility}.
\member{w0} and \member{w1} define the weighting factors
$\omega^{(0)}$ and
$\omega^{(1)}$, respectively (see equation~\ref{EQU:REG:1}).
By default \member{w0}=\None and \member{w1}=1 are used.
\member{k_at_depth} sets the value for susceptibility at depth. This is only used if susceptibility is fixed below a certain depth,
see \class{Domain Builder} in Section~\ref{Chp:ref:domain builder}.
\end{methoddesc}
\begin{methoddesc}[MagneticInversion]{setInitialGuess}{\optional{k=\None}}
sets an initial guess for the susceptibility anomaly. By default zero is used.
\end{methoddesc}
\subsection{Gravity and Magnetic Joint Inversion Driver}
For examples of usage please see Chapter~\ref{Chp:cook:joint inversion}.
\begin{classdesc}{JointGravityMagneticInversion}{\optional{solverclass=None}
, \optional{self_demagnetization=\False}
}
Driver class to perform a joint inversion of Gravity (Bouguer) and magnetic anomaly data.
This class is a sub-class of \class{InversionDriver}.
The class uses the standard \class{Regularization} for two level set functions
with cross-gradient correlation, see Chapter~\ref{Chp:ref:regularization},
\class{DensityMapping} and \class{SusceptibilityMapping} mappings, see Section~\ref{Chp:ref:mapping}, the
gravity forward model \class{GravityModel}, see Section~\ref{sec:forward gravity}
and the magnetic forward model \class{SelfDemagnetizationModel} (if \member{self_demagnetization}=\True,
see Section~\ref{sec:forward self demagnetic}) or \class{MagneticModel}
(if \member{self_demagnetization}=\False, see Section~\ref{sec:forward magnetic}).
\member{solverclass} set the solver class to be used for inversion,
see Chapter~\ref{chapter:ref:Minimization}.
By default the limited-memory Broyden-Fletcher-Goldfarb-Shanno (\emph{L-BFGS})~\cite{Nocedal1980}\index{L-BFGS} solver is used.
\end{classdesc}
\begin{methoddesc}[JointGravityMagneticInversion]{fixGravityPotentialAtBottom}{\optional{status=\True}}
If \member{status} is \True the gravity potential at the bottom is set to zero. Otherwise the gravity
potential at the top is set to zero only.
\end{methoddesc}
\begin{methoddesc}[JointGravityMagneticInversion]{fixMagneticPotentialAtBottom}{\optional{status=\True}}
If \member{status} is \True the magnetic potential at the bottom is set to zero. Otherwise the magnetic
potential at the top is set to zero only.
\end{methoddesc}
\begin{methoddesc}[JointGravityMagneticInversion]{setup}{
domainbuilder
\optional{, rho0=\None}
\optional{, drho=\None}
\optional{, rho_z0=\None}
\optional{, rho_beta=\None}
\optional{, k0=\None}
\optional{, dk=\None}
\optional{, k_z0=\None}
\optional{, k_beta=\None}
\optional{, w0=\None}
\optional{, w1=\None}
\optional{, w_gc=\None}
\optional{, rho_at_depth=\None}
\optional{, k_at_depth=\None}
}
sets up the inversion from an instance \member{domainbuilder} of a \class{DomainBuilder}, see Section~\ref{Chp:ref:domain builder}.
Gravity and magnetic data attached to the \member{domainbuilder} are considered in the inversion.
\member{rho0} defines a reference density anomaly (default is $0$),
\member{drho} defines a density anomaly (default is $2750 \frac{kg}{m^3}$),
\member{rho_z0} defines the depth weighting reference depth for density (default is \None), and
\member{rho_beta} defines the depth weighting exponent for density (default is \None),
see \class{DensityMapping} in Section~\ref{Chp:ref:mapping density}.
\member{k0} defines a reference susceptibility anomaly (default is $0$),
\member{dk} defines a susceptibility anomaly scale (default is $1$),
\member{k_z0} defines the depth weighting reference depth for susceptibility (default is \None), and
\member{k_beta} defines the depth weighting exponent for susceptibility (default is \None),
see \class{SusceptibilityMapping} in Section~\ref{Chp:ref:mapping susceptibility}.
\member{w0} and \member{w1} define the weighting factors
$\omega^{(0)}$ and
$\omega^{(1)}$, respectively (see Equation~\ref{EQU:REG:1}).
\member{w_gc} sets the weighting factor $\omega^{(c)}$ for the cross gradient term.
By default \member{w0}=\None, \member{w1}=1 and \member{w_gc}=1 are used.
\member{k_at_depth} sets the value for susceptibility at depth. This is only used if susceptibility is fixed below a certain depth,
see \class{Domain Builder} in Section~\ref{Chp:ref:domain builder}.
\member{rho_at_depth} sets the value for density at depth. This is only used if density is fixed below a certain depth,
see \class{Domain Builder} in Section~\ref{Chp:ref:domain builder}.
\end{methoddesc}
\begin{methoddesc}[JointGravityMagneticInversion]{setInitialGuess}{\optional{rho=\None, } \optional{k=\None}}
sets initial guesses for density and susceptibility anomaly.
By default zero is used for both.
\end{methoddesc}
|
