-
Notifications
You must be signed in to change notification settings - Fork 12
/
paper_pictures_demo_ICML18.py
238 lines (207 loc) · 9.36 KB
/
paper_pictures_demo_ICML18.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
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Thu Feb 1 18:47:25 2018
@author: jeremias
Description: Extract pictures
"""
import pickle
import numpy as np
from Evaluation_tool import EvaluationTool
from matplotlib import pyplot as plt
#only needed if you want to generate demo data yourself
from cp_probability_model import CpModel
from detector import Detector
import synthetic_simulations_prepare_data
import synthetic_simulations_prepare_models
#ensure that we have type 1 fonts (for ICML publishing guiedlines)
matplotlib.rcParams['pdf.fonttype'] = 42
matplotlib.rcParams['ps.fonttype'] = 42
baseline_working_directory = ("//Users//jeremiasknoblauch//Documents////OxWaSP//BOCPDMS//Code//" +
"SpatialBOCD//Paper//demo")
"""####################################################"""
"""DATA SET 1: Synthetic data set: VAR_infinity, fct. 3"""
"""####################################################"""
process_new = False #only if you want to re-run everything
"""get the right size"""
T, S1, S2 = 500, 5, 5
"""only do this if you want to generate data yourself"""
if process_new:
"""Call the relevant builder fct and add to all simulators"""
full_fct_name = "generate_VAR_infty_3"
method_to_call = getattr(synthetic_simulations_prepare_data,
full_fct_name)
sim = method_to_call(S1=S1, S2=S2, T=T)
data = sim.generate_data()
else:
"""read in data"""
data_path = baseline_working_directory
data_file = open(data_path + "//data_demo.txt", 'rb')
data = pickle.load(data_file)
data_file.close()
"""normalize"""
data = ((data - np.mean(data, axis=0))/
np.sqrt(np.var(data,axis=0)))
"""shorten to interior """
data = data[:,1:S1-1, 1:S2-1]
T = data.shape[0]
S1 = S1 - 2
S2 = S2 - 2
if process_new:
"""priors"""
intensity, a, b, b_default = 1500.0, 15.0, 0.05, 0.05
mult = 1.0
lower_AR, upper_AR = 1.0, int(mult*pow(float(T)/np.log(T), 0.25) + 1.0)
upper_VAR, lower_VAR = 1.0, int(mult*pow(float(T)/np.log(T), 1.0/6.0) + 1)
cp_model = CpModel(intensity)
hyperpar_opt = "online"
prior_var_beta_scale = 10.0
"""models"""
method_to_call = getattr(
synthetic_simulations_prepare_models,
"generate_VAR0_nbh_models")
VAR0_list = method_to_call(S1,S2,a,b_default,
prior_var_beta_scale, lower_AR, upper_AR,
hyperpar_opt)
method_to_call = getattr(
synthetic_simulations_prepare_models,
"generate_VAR4_nbh_models")
VAR4_list = method_to_call(S1,S2,a,b_default,
prior_var_beta_scale, lower_VAR, upper_VAR,
hyperpar_opt)
method_to_call = getattr(
synthetic_simulations_prepare_models,
"generate_VAR8_nbh_models")
VAR8_list = method_to_call(S1,S2,a,b_default,
prior_var_beta_scale, lower_VAR, upper_VAR,
hyperpar_opt)
model_universe = VAR0_list + VAR4_list + VAR8_list
"""detector"""
model_universe = np.array(model_universe)
model_prior = np.array(
[1.0/int(len(model_universe))] *
int(len(model_universe)))
detector = Detector(
data, model_universe, model_prior, cp_model,
S1, S2, T, exo_data=None, num_exo_vars=None,
threshold=200,
store_rl=True, store_mrl=True)#,
detector.run()
"""build evaluation tool"""
EvT = EvaluationTool()
EvT.build_EvaluationTool_via_run_detector(detector)
else:
"""Directly read results"""
result_path = baseline_working_directory
EvT = EvaluationTool()
EvT.build_EvaluationTool_via_results(result_path + "//" + "results_demo.txt")
"""Plot in panels: Raw data (offsetting the mean),
CPs + RLD
1-step-ahead-prediction + variance
model posterior"""
custom_colors_models = ['green', 'purple', 'orange', 'blue', 'darkgray']
custom_colors_series = ['black']*4
custom_linestyles =['solid']*5
offsets = np.linspace(1,S1*S2, S1*S2).reshape(S1, S2)
data_original = data.copy()
data = data + offsets*2.5
data = data.reshape(T, S1*S2)
""""Get some quantities needed for rest"""
segmentation = EvT.results[EvT.names.index("MAP CPs")][-2]
model_labels = EvT.results[EvT.names.index("model labels")]
num_models = len(np.union1d(model_labels, model_labels))
relevant_models = np.union1d([seg[1] for seg in segmentation],[seg[1] for seg in segmentation])
#exclude 7, as it just obscures the view
relevant_models = [1,2,5]
all_models = np.linspace(0, num_models-1, num_models, dtype=int)
"""detemine if you want separate model posterior plots"""
one_model_posterior_plot = True
if one_model_posterior_plot:
#paper: height_ratio = [6,6,5,24] poster: height_ratio = [5,5,5,12]
height_ratio = [6,6,5,24]
num_subplots = 4
else:
height_ratio = [8,8, 4,4,4, 18]
num_subplots = 6
"""set ylabel position"""
#paper: ylabel_coords = [-0.09, 0.5] poster: ylabel_coords = [-0.045, 0.5]
ylabel_coords = [-0.09, 0.5]
yticks_fontsize = 10
#paper: figsize argument left out poster: figsize = (12,5)
fig, ax_array = plt.subplots(num_subplots, sharex = True,
gridspec_kw = {'height_ratios':height_ratio})#,
#figsize = (12,5))
plt.subplots_adjust(hspace = .185, left = None, bottom = None, right = None, top = None)
"""plot raw data"""
fig_1 = EvT.plot_raw_TS(data, indices=[0,2,4],
custom_colors_series = custom_colors_series,
ax = ax_array[0],xlab=None,
ylabel_coords = ylabel_coords) #ax_array[0])
"""Plot + save pics (4) prediction + error"""
fig_2 = EvT.plot_prediction_error(show_var= True, data = data_original,
indices = [0],
custom_colors = ['black', 'darkgray'],
ax=ax_array[1],
time_range = np.linspace(1,T-2,T-2,dtype=int),
aspect_ratio = 'auto', xlab = None, ylab = "PE",
ylabel_coords = ylabel_coords)
"""Plot + save pictures: (3) model posteriors"""
"""Plot + save pictures: (2) Raw TS with MAP CPs"""
#backconverter = tick/C2-C1 (backconv+C1)*C2 = tick
C1, C2 = 100, 1
#raw_ticks = [pow(10, -18), pow(10, -15), pow(10, -13), pow(10, -11)]
colorbar_ticks = None #[C2*(r+C1) for r in raw_ticks]
fig_4 = EvT.plot_run_length_distr(mark_max=True,upper_limit=250,
aspect_ratio='auto', C1=C1, C2=C2,
time_range = np.linspace(1,T-2,T-2, dtype=int),
CP_legend = True, CP_legend_fontsize = 8,
CP_custom_legend_labels = ["AR(2)", "AR(3)", "VAR4(2)"],
additional_legend_labels = ["VAR8(1)"],
additional_legend_colors = ["blue"],
CP_exclude_indices = [7], #just obscures view
custom_colors = custom_colors_models,
custom_linestyles =custom_linestyles,
ax = ax_array[-1], figure = fig, space_to_colorbar = 0.275,
orientation = 'horizontal',
xlab=None, colorbar_location = "bottom",
ylabel_coords = ylabel_coords,
colorbar_ticks_num = 1)
if one_model_posterior_plot:
#custom_colors = ['purple', 'orange', 'red', 'green']
fig_3 = EvT.plot_model_posterior(indices= [1,2,5,6],#relevant_models.append(30),
plot_type = "trace",
y_axis_labels = ["mod1", "mod2", "mod3"],
log_format=False, aspect = 'auto',
show_MAP_CPs = False,
custom_colors = custom_colors_models,
ax = ax_array[2], xlab = None,
ylabel_coords = ylabel_coords)
# if one_model_posterior_plot:
##custom_colors = ['purple', 'orange', 'red', 'green']
# fig_3 = EvT.plot_model_posterior(indices=relevant_models,
# plot_type = "MAP",
# y_axis_labels = ["mod1", "mod2", "mod3"],
# log_format=False, aspect = 'auto',
# show_MAP_CPs = False,
# custom_colors = custom_colors_models,
# ax = None, xlab = None)
else:
for ind, mod in zip(range(0, len(relevant_models)), relevant_models):
# if ind == 2:
# xlab = "Time"
# else:
xlab = None
if ind == 1:
ylab = "P(m|y)"
else:
ylab = None
fig_3 = EvT.plot_model_posterior(indices=[relevant_models[ind]],
log_format=False, aspect = 'auto',
show_MAP_CPs = False,
custom_colors = [custom_colors_models[ind]],
ax = ax_array[2+ind], xlab = xlab, ylab = ylab,
ylabel_coords = ylabel_coords)
#plt.show()
fig.savefig(baseline_working_directory + "//demo_picture.pdf",
format = "pdf", orientation = "portrait",
dpi = 800)