-
-
Notifications
You must be signed in to change notification settings - Fork 231
/
input_manipulation.py
225 lines (170 loc) · 8.15 KB
/
input_manipulation.py
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
"""
This file is part of gempy.
gempy is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
gempy 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 General Public License for more details.
You should have received a copy of the GNU General Public License
along with gempy. If not, see <http://www.gnu.org/licenses/>.
Tested on Ubuntu 16
Created on 23/06/2018
@author: Miguel de la Varga, Alexander Schaaf
"""
import numpy as np
import pandas as pn
def find_surface_points_from_block(block, value):
"""
TODO: Is this dep by find_interfaces_from_block_bottoms?
Find the voxel at an interface. We shift left since gempy is based on bottoms
Args:
block (ndarray):
value:
Returns:
"""
A = block > value
# Matrix shifting along axis
B = A #
x_shift = B[:-1, :, :] ^ B[1:, :, :]
# Matrix shifting along axis
y_shift = B[:, :-1, :] ^ B[:, 1:, :]
# Matrix shifting along axis
z_shift = B[:, :, :-1] ^ B[:, :, 1:]
final_bool = np.zeros_like(block, dtype=bool)
final_bool[:-1, :-1, :-1] = x_shift[:, :-1, :-1] + y_shift[:-1, :, :-1] + z_shift[-1:, -1:, :]
return final_bool
def find_interfaces_from_block_bottoms(block, value, shift=2):
"""
Find the voxel at an interface. We shift left since gempy is based on bottoms
Args:
block (ndarray): matrix with the scalar values
value: value of which you are looking the interfaces
shift (int): Number of elements shifted
Returns:
"""
A = block == value
final_bool = np.zeros_like(block, dtype=bool)
# Matrix shifting along axis 0
x_shift = A[:-shift, :, :] ^ A[shift:, :, :]
# Matrix shifting along axis 1
y_shift = A[:, :-shift, :] ^ A[:, shift:, :]
# Matrix shifting along axis 2
z_shift = A[:, :, :-shift] ^ A[:, :, shift:]
final_bool[shift:, shift:, shift:] = (x_shift[:, shift:, shift:] +
y_shift[shift:, :, shift:] +
z_shift[shift:, shift:, :])
return final_bool
def surface_points_from_surface_points_block(block_bool, block_grid, formation='default_formation', series='Default_series',
formation_number=1, order_series=1, n_points=20):
assert np.ravel(block_bool).shape[0] == block_grid.shape[0], 'Grid and block block must have the same size. If you' \
'are importing a model from noddy make sure that the' \
'resolution is the same'
coord_select = block_grid[np.ravel(block_bool)]
loc_points = np.linspace(0, coord_select.shape[0]-1, n_points, dtype=int)
# Init dataframe
p = pn.DataFrame(columns=['X', 'Y', 'Z', 'formation', 'series', 'formation_number',
'order_series', 'isFault'])
p[['X', 'Y', 'Z']] = pn.DataFrame(coord_select[loc_points])
p['formation'] = formation
p['series'] = series
p['formation_number'] = formation_number
p['order_series'] = order_series
return p
def set_surface_points_from_block(geo_data, block, block_grid=None, n_points=20, reset_index=False):
values = np.unique(np.round(block))
values.sort()
values = values[:-1]
if block_grid is None:
block_grid = geo_data._grid.values
for e, value in enumerate(values):
block_bool = find_surface_points_from_block(block, value)
geo_data.set_interface_object(surface_points_from_surface_points_block(block_bool, block_grid,
formation='formation_'+str(e), series='Default_series',
formation_number=e, order_series=1,
n_points=n_points), append=True)
if reset_index:
geo_data._surface_points.reset_index(drop=True, inplace=True)
return geo_data
class VanMisesFisher:
def __init__(self, mu, kappa, dim=3):
"""van Mises-Fisher distribution for sampling vector components from n-dimensional spheres.
Adapted from source: https://github.com/pymc-devs/pymc3/issues/2458
a von Mises-Fisher distribution can be fit using scipy.stats.vonmises
Args:
mu (np.ndarray): Mean direction of vector [Gx, Gy, Gz]
kappa (float): Concentration parameter (the lower the higher the spread on the sphere)
dim (int, optional): Dimensionality of the Sphere
"""
self.mu = mu
self.kappa = kappa
self.dim = dim
def rvs(self, n=1):
"""Obtain n samples from van Mises-Fisher distribution.
Args:
n (int): Number of samples to draw
Returns:
np.ndarray with shape (n, 3) containing samples.
"""
result = np.zeros((n, self.dim))
for nn in range(n):
# sample offset from center (on sphere) with spread kappa
w = self._sample_weight()
# sample a point v on the unit sphere that's orthogonal to mu
v = self._sample_orthonormal_to()
# compute new point
result[nn, :] = v * np.sqrt(1. - w** 2) + w * self.mu
return result
def _sample_weight(self):
"""Who likes documentation anyways. This is totally intuitive and trivial."""
dim = self.dim - 1 # since S^{n-1}
b = dim / (np.sqrt(4. * self.kappa ** 2 + dim ** 2) + 2 * self.kappa)
x = (1. - b) / (1. + b)
c = self.kappa * x + dim * np.log(1 - x ** 2)
while True:
z = np.random.beta(dim / 2., dim / 2.)
w = (1. - (1. + b) * z) / (1. - (1. - b) * z)
u = np.random.uniform(low=0, high=1)
if self.kappa * w + dim * np.log(1. - x * w) - c >= np.log(u):
return w
def _sample_orthonormal_to(self):
"""Who likes documentation anyways. This is totally intuitive and trivial."""
v = np.random.randn(self.mu.shape[0])
proj_mu_v = self.mu * np.dot(self.mu, v) / np.linalg.norm(self.mu)
orthto = v - proj_mu_v
return orthto / np.linalg.norm(orthto)
def stats(self):
return self.mu, self.kappa
def change_data(interp_data, geo_data_stoch, priors):
"""Changes input data with prior distributions (scipy.stats distributions) given in list.
Prior distribution objects must contain .rvs() method for drawing samples.
Args:
interp_data:
geo_data_stoch:
priors:
verbose:
Returns:
"""
prior_draws = []
for prior in priors:
if hasattr(prior, "gradient"):
value = prior.rvs()
else:
value = prior.rvs() / interp_data.rescaling_factor
prior_draws.append(value)
if prior.index_interf is not None:
if prior.replace: # replace the value
# geo_data.interfaces.set_value(prior.index_interf, prior.column, prior.rvs() / rf)
interp_data.geo_data_res.interfaces.loc[prior.index_interf, prior.column] = value
else: # add value
interp_data.geo_data_res.interfaces.loc[prior.index_interf, prior.column] = geo_data_stoch.interfaces.loc[
prior.index_interf, prior.column] + value
if prior.index_orient is not None:
if prior.replace: # replace the value
interp_data.geo_data_res._orientations.loc[prior.index_orient, prior.column] = value
else: # add value
interp_data.geo_data_res._orientations.loc[prior.index_orient, prior.column] = geo_data_stoch._orientations.loc[
prior.index_orient, prior.column] + value
return prior_draws