-
Notifications
You must be signed in to change notification settings - Fork 2
/
Copy pathdataset_2_formatter.py
239 lines (197 loc) · 11.2 KB
/
dataset_2_formatter.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
226
227
228
229
230
231
232
233
234
235
236
237
238
239
from preprocessing import *
# Data set 9: Applying forces to 100 nodes at a time. Forces have 30 different magnitudes and 15 different directions.
def dataset_2(idx):
input_path = 'Data_Generator/code/input_2/'
output_path = 'Data_Generator/code/output_2/csv/'
# Initialization
elements_filename = input_path + 'elements.csv'
element_id_filename = input_path + 'element_ID.csv'
xyz_filename = input_path + 'xyz.csv'
support_nodes_filename = input_path + 'bcSupportList.csv'
pres_nodes_filename = input_path + 'bcPrescribeList.csv'
boundary_faces_filename = input_path + 'boundary_faces.csv'
boundary_id_filename = input_path + 'boundary_marker.csv'
num_nodes = 9118 # number of nodes per each graph
num_p_nodes = 100
num_dirs = 15
t_steps = 30
f_magnitude = 20 # units: Newtons
if idx == 1:
formatted_data_path = 'dataset_2/a/'
filename_force_dir_x = input_path + 'force_dir_x_a.csv'
filename_force_dir_y = input_path + 'force_dir_y_a.csv'
filename_force_dir_z = input_path + 'force_dir_z_a.csv'
offset_output = 0
elif idx == 2:
formatted_data_path = 'dataset_2/b/'
filename_force_dir_x = input_path + 'force_dir_x_b.csv'
filename_force_dir_y = input_path + 'force_dir_y_b.csv'
filename_force_dir_z = input_path + 'force_dir_z_b.csv'
offset_output = 15
elif idx == 3:
formatted_data_path = 'dataset_2/c/'
filename_force_dir_x = input_path + 'force_dir_x_c.csv'
filename_force_dir_y = input_path + 'force_dir_y_c.csv'
filename_force_dir_z = input_path + 'force_dir_z_c.csv'
offset_output = 30
elif idx == 4:
formatted_data_path = 'dataset_2/d/'
filename_force_dir_x = input_path + 'force_dir_x_d.csv'
filename_force_dir_y = input_path + 'force_dir_y_d.csv'
filename_force_dir_z = input_path + 'force_dir_z_d.csv'
offset_output = 45
elif idx == 5:
formatted_data_path = 'dataset_2/e/'
filename_force_dir_x = input_path + 'force_dir_x_e.csv'
filename_force_dir_y = input_path + 'force_dir_y_e.csv'
filename_force_dir_z = input_path + 'force_dir_z_e.csv'
offset_output = 60
elif idx == 6:
formatted_data_path = 'dataset_2/f/'
filename_force_dir_x = input_path + 'force_dir_x_f.csv'
filename_force_dir_y = input_path + 'force_dir_y_f.csv'
filename_force_dir_z = input_path + 'force_dir_z_f.csv'
offset_output = 75
elif idx == 7:
formatted_data_path = 'dataset_2/g/'
filename_force_dir_x = input_path + 'force_dir_x_g.csv'
filename_force_dir_y = input_path + 'force_dir_y_g.csv'
filename_force_dir_z = input_path + 'force_dir_z_g.csv'
offset_output = 90
elif idx == 8:
formatted_data_path = 'dataset_2/h/'
filename_force_dir_x = input_path + 'force_dir_x_h.csv'
filename_force_dir_y = input_path + 'force_dir_y_h.csv'
filename_force_dir_z = input_path + 'force_dir_z_h.csv'
offset_output = 105
elif idx == 9:
formatted_data_path = 'dataset_2/i/'
filename_force_dir_x = input_path + 'force_dir_x_i.csv'
filename_force_dir_y = input_path + 'force_dir_y_i.csv'
filename_force_dir_z = input_path + 'force_dir_z_i.csv'
offset_output = 120
elif idx == 10:
formatted_data_path = 'dataset_2/j/'
filename_force_dir_x = input_path + 'force_dir_x_j.csv'
filename_force_dir_y = input_path + 'force_dir_y_j.csv'
filename_force_dir_z = input_path + 'force_dir_z_j.csv'
offset_output = 135
elif idx == 11:
formatted_data_path = 'dataset_2/h/'
filename_force_dir_x = input_path + 'force_dir_x_k.csv'
filename_force_dir_y = input_path + 'force_dir_y_k.csv'
filename_force_dir_z = input_path + 'force_dir_z_k.csv'
offset_output = 150
else:
raise ValueError('Dataset for offsets above 11 do not exist.')
######################################################################################################
# Output file names: 1) formatted input files, 2) intermediate files, 3) formatted output files
# ** We only need to upload the formatted input and output files to Google Drive
# 1) formatted input files
adj_matrix_filename = formatted_data_path + 'A.csv'
node_att_filename = formatted_data_path + 'node_attributes.csv'
graph_indicator_filename = formatted_data_path + 'graph_indicator.csv'
graph_labels_filename = formatted_data_path + 'graph_labels.csv'
# -----------------------------------------------------------------------------
# 2) intermediate files
node_material_id_filename = formatted_data_path + 'node_material_id.csv'
rigid_nodes_filename = formatted_data_path + 'rigid_nodes_id.csv'
adj_matrix_short_filename = formatted_data_path + 'A_partial.csv'
# -----------------------------------------------------------------------------
# 3) formatted output files
output_displacement_filename = formatted_data_path + 'output_displacement.csv'
###################################################################################################################
# Creating the adjacency matrix
# A = adj_matrix_builder(elements_filename, num_nodes)
# adj_matrix_full_format(A, num_nodes, num_dirs, t_steps, adj_matrix_filename)
###################################################################################################################
# Creating the feature matrix
df_xyz = xyz(xyz_filename)
df_mat_id = node_material_assign(elements_filename, element_id_filename, boundary_faces_filename, boundary_id_filename,
num_nodes, node_material_id_filename)
df_rigid_id = node_support_assign(support_nodes_filename, num_nodes, rigid_nodes_filename)
df_features = feature_constructor_2(num_nodes, df_xyz, df_mat_id, df_rigid_id, pres_nodes_filename,
filename_force_dir_x, filename_force_dir_y, filename_force_dir_z, f_magnitude, t_steps)
df_features.to_csv(node_att_filename, encoding='utf-8', header=False, index=False)
###################################################################################################################
# Creating the Graph indicator file (which determines which node belongs to which graph)
node_graph_labels = graph_indicator_2(num_nodes, num_dirs, t_steps)
node_graph_labels.to_csv(graph_indicator_filename, header=False, index=False)
# ###################################################################################################################
#
# # Graph labels
graph_labels = graph_label_2(num_dirs, t_steps)
graph_labels.to_csv(graph_labels_filename, header=False, index=False)
# ###################################################################################################################
#
# Output for 100 load nodes, 15 directions, and 30 time steps. Force is applied to all nodes at once.
output_files = []
for i in range(1, num_dirs + 1):
s1 = "displacement_nodes_dir_"
s2 = str(i + offset_output)
s3 = ".csv"
s = output_path + s1 + s2 + s3
print(s)
output_files.append(s)
df_1 = output_format(output_files[0], t_steps)
df_2 = output_format(output_files[1], t_steps)
df_3 = output_format(output_files[2], t_steps)
df_4 = output_format(output_files[3], t_steps)
df_5 = output_format(output_files[4], t_steps)
df_6 = output_format(output_files[5], t_steps)
df_7 = output_format(output_files[6], t_steps)
df_8 = output_format(output_files[7], t_steps)
df_9 = output_format(output_files[8], t_steps)
df_10 = output_format(output_files[9], t_steps)
df_11 = output_format(output_files[10], t_steps)
df_12 = output_format(output_files[11], t_steps)
df_13 = output_format(output_files[12], t_steps)
df_14 = output_format(output_files[13], t_steps)
df_15 = output_format(output_files[14], t_steps)
frame = [df_1, df_2, df_3, df_4, df_5, df_6, df_7, df_8, df_9, df_10, df_11, df_12, df_13, df_14, df_15]
df_output = pd.concat(frame)
df_output.reset_index(drop=True, inplace=True)
df_output.to_csv(output_displacement_filename, header=False, index=False)
print(df_output)
# ##################################################################################################################
# ##################################################################################################################
# ##################################################################################################################
# ##################################################################################################################
# FEATURE PREPROCESSING
data_node_att = np.loadtxt(formatted_data_path + 'node_attributes.csv', delimiter=',')
# --------------------------------------------------------------------------------------------------------------------
# Node attribute format
# {x} {y} {z} {material_ID} {Rigid} {Force magnitude} {F_x} {F_y} {F_z}
# 0 1 2 3 4 5 6 7 8
# material id 1: brain, 2: tumour
# ----------------------------------------------------------------------------------------------------------------------
# ############# Continuous encoding of rigid ID and material ID ###############
physics_prop = np.zeros((data_node_att.shape[0], 1), dtype=float)
for j in range(data_node_att.shape[0]):
if data_node_att[j, 4] == 1:
physics_prop[j, 0] = 0
else:
if data_node_att[j, 3] == 1:
physics_prop[j, 0] = 1
else:
physics_prop[j, 0] = 0.4
# ############################################################################
# ############# Multiplication of force magnitude by direction ###############
force_magnitude = data_node_att[:, 5]
x_mag_and_direction = np.multiply(data_node_att[:, 6], force_magnitude)
y_mag_and_direction = np.multiply(data_node_att[:, 7], force_magnitude)
z_mag_and_direction = np.multiply(data_node_att[:, 8], force_magnitude)
# ----------------------------------------------------------------------------------------------------------------------
x_mag_and_direction = np.reshape(x_mag_and_direction, (-1, 1))
y_mag_and_direction = np.reshape(y_mag_and_direction, (-1, 1))
z_mag_and_direction = np.reshape(z_mag_and_direction, (-1, 1))
# ----------------------------------------------------------------------------------------------------------------------
feature_normalized = np.concatenate((data_node_att[:, 0:3], physics_prop, x_mag_and_direction, y_mag_and_direction,
z_mag_and_direction), axis=1)
# ----------------------------------------------------------------------------------------------------------------------
print(feature_normalized)
print(feature_normalized.shape)
r = np.ptp(feature_normalized, axis=0)
print(r)
np.savetxt(formatted_data_path + "/node_attributes_raw.csv", feature_normalized, delimiter=",",
fmt=('%f, %f, %f, %f, %f, %f, %f'))