/
load_experimental_data.py
261 lines (206 loc) · 10.6 KB
/
load_experimental_data.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
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
'''
Load experimental data in python, because Matlab sucks ass.
'''
import sys
import matplotlib.pyplot as plt
# import matplotlib.patches as plt_patches
# import matplotlib.gridspec as plt_grid
import os
import os.path
# import bottleneck as bn
import utils
from plots_experimental_data import *
from experimentalloaderbays09 import *
from experimentalloaderdualrecall import *
from experimentalloadergorgo11seq import *
from experimentalloadergorgo11sim import *
######
def load_data_simult(data_dir='../../experimental_data/',
fit_mixture_model=False,
compute_bootstrap=False):
'''
Convenience function, automatically load the Gorgoraptis_2011 dataset.
'''
if data_dir == '../../experimental_data/':
experim_datadir = os.environ.get('WORKDIR_DROP',
os.path.split(utils.__file__)[0])
data_dir = os.path.normpath(os.path.join(experim_datadir, data_dir))
expLoader = ExperimentalLoaderGorgo11Simultaneous(
dict(
name='gorgo11',
filename='Exp2_withcolours.mat',
datadir=os.path.join(data_dir, 'Gorgoraptis_2011'),
parameters=dict(
fit_mixture_model=fit_mixture_model,
mixture_model_cache='em_gorgo_sim_basicmodel.pickle',
collapsed_mixture_model_cache='collapsed_em_gorgo_sim_new.pickle',
should_compute_bootstrap=compute_bootstrap,
bootstrap_cache='bootstrap_gorgo_500samples.pickle',
nb_bootstrap_samples=500
)))
return expLoader.dataset
def load_data_gorgo11(data_dir='../../experimental_data/',
fit_mixture_model=False,
compute_bootstrap=False):
'''
Convenience function, automatically load the Gorgo11 simultaneous dataset.
'''
return load_data_simult(data_dir, fit_mixture_model)
def load_data_gorgo11_sequential(data_dir='../../experimental_data/',
fit_mixture_model=False):
'''
Convenience function, automatically load the Gorgo11 sequential dataset.
'''
if data_dir == '../../experimental_data/':
experim_datadir = os.environ.get('WORKDIR_DROP',
os.path.split(utils.__file__)[0])
data_dir = os.path.normpath(os.path.join(experim_datadir, data_dir))
expLoader = ExperimentalLoaderGorgo11Sequential(
dict(
name='gorgo11seq',
filename='Exp1.mat',
datadir=os.path.join(data_dir, 'Gorgoraptis_2011'),
parameters=dict(
fit_mixture_model=fit_mixture_model,
mixture_model_cache='em_gorgo_seq_basicmodel.pickle',
collapsed_mixture_model_cache='collapsed_em_gorgo_seq_new.pickle'
)))
return expLoader.dataset
def load_data_bays09(data_dir='../../experimental_data/',
fit_mixture_model=False,
compute_bootstrap=False):
'''
Convenience function, automatically load the Bays2009 dataset.
'''
if data_dir == '../../experimental_data/':
experim_datadir = os.environ.get('WORKDIR_DROP',
os.path.split(utils.__file__)[0])
data_dir = os.path.normpath(os.path.join(experim_datadir, data_dir))
expLoader = ExperimentalLoaderBays09(
dict(
name='bays09',
filename='colour_data.mat',
datadir=os.path.join(data_dir, 'Bays2009'),
parameters=dict(
fit_mixture_model=fit_mixture_model,
mixture_model_cache='em_bays_basicmodel.pickle',
collapsed_mixture_model_cache='collapsed_em_bays_new.pickle',
should_compute_bootstrap=compute_bootstrap,
bootstrap_cache='bootstrap_bays_500samples.pickle',
nb_bootstrap_samples=1000)))
return expLoader.dataset
def load_data_dualrecall(data_dir='../../experimental_data/',
fit_mixture_model=False):
'''
Convenience function, automatically load the Double recall dataset (unpublished).
'''
if data_dir == '../../experimental_data/':
experim_datadir = os.environ.get('WORKDIR_DROP',
os.path.split(utils.__file__)[0])
data_dir = os.path.normpath(os.path.join(experim_datadir, data_dir))
expLoader = ExperimentalLoaderDualRecall(
dict(
name='dualrecall',
filename='rate_data.mat',
datadir=os.path.join(data_dir, 'DualRecall_Bays'),
parameters=dict(
fit_mixture_model=fit_mixture_model,
mixture_model_cache='em_dualrecall_allitems.pickle',
collapsed_mixture_model_cache='collapsed_em_dualrecall.pickle')))
return expLoader.dataset
def load_data(experiment_id='bays09',
data_dir='../../experimental_data/',
fit_mixture_model=True,
compute_bootstrap=False):
'''
Load the appropriate dataset given an experiment_id.
'''
if experiment_id == 'bays09':
return load_data_bays09(
data_dir=data_dir,
fit_mixture_model=fit_mixture_model,
compute_bootstrap=compute_bootstrap)
elif experiment_id == 'gorgo11':
return load_data_gorgo11(
data_dir=data_dir,
fit_mixture_model=fit_mixture_model,
compute_bootstrap=compute_bootstrap)
elif experiment_id == 'gorgo11_sequential':
return load_data_gorgo11_sequential(
data_dir=data_dir, fit_mixture_model=fit_mixture_model)
elif experiment_id == 'dualrecall':
return load_data_dualrecall(
data_dir=data_dir, fit_mixture_model=fit_mixture_model)
else:
raise ValueError('Experiment_id %s unknown.' % experiment_id)
if __name__ == '__main__':
## Load data
experim_datadir = os.environ.get('WORKDIR_DROP',
os.path.split(utils.__file__)[0])
data_dir = os.path.normpath(
os.path.join(experim_datadir, '../../experimental_data/'))
# data_dir = '/Users/loicmatthey/Dropbox/UCL/1-phd/Work/Visual_working_memory/experimental_data/'
# data_dir = os.path.normpath(os.path.join(experim_datadir, '../experimental_data/'))
print sys.argv
if True or (len(sys.argv) > 1 and sys.argv[1]):
# keys:
# 'probe', 'delayed', 'item_colour', 'probe_colour', 'item_angle', 'error', 'probe_angle', 'n_items', 'response', 'subject']
# (data_sequen, data_simult, data_dualrecall) = load_multiple_datasets([dict(filename='Exp1.mat', parameters=dict(datadir=os.path.join(data_dir, 'Gorgoraptis_2011'))), dict(filename='Exp2_withcolours.mat', parameters=dict(datadir=os.path.join(data_dir, 'Gorgoraptis_2011'), fit_mixture_model=True)), dict(filename=os.path.join(data_dir, 'DualRecall_Bays', 'rate_data.mat'), parameters=dict(fit_mixture_model=True))])
# (data_simult,) = load_multiple_datasets([dict(name='Gorgo_simult', filename='Exp2_withcolours.mat', parameters=dict(datadir=os.path.join(data_dir, 'Gorgoraptis_2011'), fit_mixture_model=True, mixture_model_cache='em_simult.pickle'))])
# (data_bays2009, ) = load_multiple_datasets([dict(name='Bays2009', filename='colour_data.mat', parameters=dict(datadir=os.path.join(data_dir, 'Bays2009'), fit_mixture_model=True, mixture_model_cache='em_bays.pickle', should_compute_bootstrap=True, bootstrap_cache='bootstrap_1000samples.pickle'))])
# data_bays2009 = load_data_bays09(data_dir=data_dir, fit_mixture_model=True)
# data_gorgo11 = load_data_gorgo11(data_dir=data_dir, fit_mixture_model=True)
# data_dualrecall = load_data_dualrecall(data_dir=data_dir, fit_mixture_model=True)
# data_gorgo11_sequ = load_data_gorgo11_sequential(data_dir=data_dir, fit_mixture_model=True)
data_bays09 = load_data('bays09', data_dir=data_dir)
# Check for bias towards 0 for the error between response and all items
# check_bias_all(data_simult)
# Check for bias for the best non-probe
# check_bias_bestnontarget(data_simult)
# check_bias_all(data_sequen)
# check_bias_bestnontarget(data_sequen)
# print data_simult['precision_subject_nitems_bays']
# print data_simult['precision_subject_nitems_theo']
# prec_exp = np.mean(data_simult['precision_subject_nitems_bays'], axis=0)
# prec_theo = np.mean(data_simult['precision_subject_nitems_theo'], axis=0)
# fi_fromexp = prec_exp**2./4.
# fi_fromtheo = prec_theo**2./4.
# print "Precision experim", prec_exp
# print "FI from exp", fi_fromexp
# print "Precision no chance level removed", prec_theo
# print "FI no chance", fi_fromtheo
# plots_check_oblique_effect(data_simult, nb_bins=50)
# np.save('processed_experimental_230613.npy', dict(data_simult=data_simult, data_sequen=data_sequen))
# plots_dualrecall(data_dualrecall)
plt.rcParams['font.size'] = 16
dataio = None
# dataio = DataIO.DataIO(label='experiments_bays2009')
# plots_check_bias_nontarget(data_simult, dataio=dataio)
# plots_check_bias_bestnontarget(data_simult, dataio=dataio)
# plots_check_bias_nontarget_randomized(data_simult, dataio=dataio)
# plots_bays2009(data_bays2009, dataio=dataio)
# dataio = DataIO.DataIO(label='experiments_gorgo11')
# plots_gorgo11(data_gorgo11, dataio)
# plots_precision(data_gorgo11, dataio)
# plots_precision(data_bays2009, dataio)
# dataio = DataIO.DataIO(label='experiments_bays2009')
# plot_bias_close_feature(data_bays2009, dataio)
# dataio = DataIO.DataIO(label='experiments_gorgo11')
# plot_bias_close_feature(data_gorgo11, dataio)
# plot_compare_bic_collapsed_mixture_model(data_bays2009, dataio)
# plot_compare_bic_collapsed_mixture_model(data_gorgo11, dataio)
if False:
for subj in data_bays2009['data_subject_split']['subjects_space'][:5]:
for nitems_i, nitems in enumerate(
data_bays2009['data_subject_split']['nitems_space']):
utils.scatter_marginals(
data_bays2009['data_subject_split']['data_subject'][subj][
'targets'][nitems_i],
data_bays2009['data_subject_split']['data_subject'][subj][
'responses'][nitems_i],
title='Subject %d, %d items' % (subj, nitems))
# dataio = DataIO.DataIO(label='experiments_gorgo11_seq')
# plots_gorgo11_sequential(data_gorgo11_sequ, dataio)
# plots_gorgo11_sequential_collapsed(data_gorgo11_sequ, dataio)
# plot_compare_bic_collapsed_mixture_model_sequential(data_gorgo11_sequ, dataio)
plt.show()