-
Notifications
You must be signed in to change notification settings - Fork 0
/
beests_model.py
215 lines (179 loc) · 7.92 KB
/
beests_model.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
from __future__ import division
import os
import sys
import cStringIO
#import ipdb
#sys.path.insert(0, '/gpfs_home/dscott3/code/beests-hack-versioned/stopsignal/stopsignal_wtf') # location of src
sys.path.insert(0, './') # location of src
sys.path.insert(0, './stopsignal_wtf') # location of src
sys.path.insert(0, './src') # location of src
if (len(sys.argv) != 2):
print("Please provide exactly one argument, the path to the analysis directory.")
print("This path must contain 'sst_data.csv' and 'analysis.txt', the BEESTS data and config files.")
print(' ')
print('You have provided:')
print(sys.argv)
exit(1)
path_analysisDir = sys.argv[1]
dataFile = sys.argv[1] + 'sst_data.csv'
path__analysisDescription = sys.argv[1] + 'analysis.txt'
print('Using analysis directory : ' + path_analysisDir)
print('Using analysis configuration: ' + path__analysisDescription)
print('Using data found in : ' + dataFile)
# Load configuration variables from analysis.txt file
config_vars = dict()
with open(path__analysisDescription) as f:
for line in f:
# Allow for some formatting in analysis.txt, which can be ignored.
if line == '\n': continue
if line[0] == '#': continue
if line[0] == ' ': continue
# Otherwise extract parameter values
eq_index = line.find(',')
var_name = line[:eq_index].strip()
var_name = var_name.replace('"', '').strip()
number = line[eq_index + 1:].strip()
number = number.replace('"', '').strip()
config_vars[var_name] = number
samples = int(config_vars["samples"])
burnIn = int(config_vars["burn-in"])
numberOfChains = int(config_vars["number of chains"])
thinning = int(config_vars["thinning"])
estimatesForAll = (config_vars["estimates for subjects or groups"]=="All")
summaryStatistics = (int(config_vars["summary statistics"]) != 0)
posteriorDistributions = (int(config_vars["posterior distributions"]) != 0)
mcmcChains = (int(config_vars["mcmc chains"]) != 0)
deviance = (int(config_vars["deviance"]) != 0)
posteriorPredictors = (int(config_vars["posterior predictors"]) != 0)
posteriorPredictorsSamples = int(config_vars["posterior predictor samples"])
numCores = int(config_vars["cpu cores"])
int_lower = int(config_vars["limits of integration lower"])
int_upper = int(config_vars["limits of integration upper"])
version = (int(config_vars["model trigger failure"]) == 0)
import stopsignal_wtf as stopsignal
import multiprocessing
from kabuki.analyze import gelman_rubin
from kabuki.utils import get_traces
from kabuki.utils import save_csv
from kabuki.utils import load_csv
import numpy
import math
from kabuki import Hierarchical
# args << "-u"
# << _programDir.absoluteFilePath("stopsignal/run.py")
# << _dataFile
# << _analysisDir
# << _analysisDescriptionPath;
data = load_csv(dataFile)
ncol = len(data.iloc[1])
actual_cores = multiprocessing.cpu_count()
if numCores < 1:
print('Warning: Inputting 0 CPU core is silly. BEESTs will use the default number of ' + str(actual_cores-1) + ' cores.')
numCores = actual_cores -1
if samples < 1:
#print "The total number of MCMC samples must be greater than zero."
sys.exit()
if samples <= burnIn:
#print "The total number of MCMC samples must be higher than the number of burn-in samples."
sys.exit()
if thinning < 1:
#print "The thinning factor must be higher than 0."
sys.exit()
if ((samples-burnIn)/thinning) < 1:
#print "No MCMC samples will be retained. Increse the number of retained samples or decrease the thinning factor."
sys.exit()
if posteriorPredictors == True:
if (((samples-burnIn)/thinning)*numberOfChains) < posteriorPredictorsSamples:
#print "The number of posterior predictive samples cannot be higher than the number of retained MCMC samples."
sys.exit()
if (int_lower >= int_upper):
sys.exit()
def check_par(pars):
for par in pars:
if (float(config_vars[par+" upper"]) <= float(config_vars[par+" lower"])):
sys.exit()
if ((float(config_vars[par+" start"]) < float(config_vars[par+" lower"])) or (float(config_vars[par+" start"]) > float(config_vars[par+" upper"]))):
sys.exit()
print('BEESTS will fit the standard model.')
#check_par(pars=("go mu", "go sigma", "go tau", "stop mu", "stop sigma", "stop tau", "go mu sd",
# "go sigma sd", "go tau sd", "stop mu sd", "stop sigma sd", "stop tau sd", "pf", "pf sd"))
#guess if rt data is in msec vs. sec
rts = data["rt"]
rts = rts[rts!=-999]
if ((min(rts)<80) | (max(rts)>2500)):
print('The maximum and/or minimum RT in your dataset is unlikely. Are you sure that your data are in milliseconds?')
print('Min RT: ')
print(min(rts))
print('Max RT: ')
print(max(rts))
#--------------------------------------------------------------Start sampling
os.chdir(path_analysisDir)
models =[]
local_models = []
def run_model(i):
ss = stopsignal.StopSignal(data)
save_stdout = sys.stdout
sys.stdout = cStringIO.StringIO()
ss.find_starting_values()
ss.sample(samples,burn=burnIn,thin=thinning,tune_throughout=False, db='pickle', dbname='remote_traces' + str(i+1) + '.db')
sys.stdout = save_stdout
return ss
if __name__ == "__main__":
if actual_cores < numCores:
if actual_cores==1:
print('Your system doesn\'t have ' + str(numCores) + ' cores. BEESTs will use 1 core.')
numCores = actual_cores
else:
print('Your system doesn\'t have ' + str(numCores) + ' cores. BEESTs will use the default number of ' + str(actual_cores-1) + ' core(s).')
numCores = actual_cores-1
else:
print('Your system has ' + str(actual_cores) + ' cores. BEESTs will use ' + str(numCores) + ' core(s)')
n_runs = math.ceil((numberOfChains/numCores))
n_runs = numpy.array(n_runs).astype('int')
num_remote = numberOfChains-n_runs
num_remote = numpy.array(num_remote).astype('int')
n_pool = numCores-1
if num_remote>0:
remote_model = []
pool = multiprocessing.Pool(processes=n_pool)
results = pool.map_async(run_model, range(num_remote))
for i in range(n_runs):
run_id = i+1
beg = run_id + (i * n_pool)
end = beg + n_pool
if i == (n_runs-1):
end = numberOfChains
ran_chains = range(beg,end+1)
print('Running chain(s) ' + str(ran_chains))
ss = stopsignal.StopSignal(data)
print('\n Computing start values. It may take a few minutes. Or hours...')
save_stdout = sys.stdout
sys.stdout = cStringIO.StringIO()
ss.find_starting_values()
sys.stdout = save_stdout
ss.sample(samples,burn=burnIn,thin=thinning,tune_throughout=False, db='pickle', dbname='local_traces' + str(i+1) + '.db')
local_models.append(ss)
if i == (n_runs-1):
print('Waiting for the other chains to finish...')
if num_remote>0:
models = results.get()
for i in range(n_runs):
models.append(local_models[i])
#print(len(models))
print "Finished sampling!"
if numberOfChains > 1:
Rhat = gelman_rubin(models)
print('\n Gelman-Rubin Rhat diagnostic:')
for key in Rhat.keys():
print((key, Rhat[key]))
name_dataFile = dataFile.replace(".csv","")
for i in range(numberOfChains):
save_csv(get_traces(models[i]), 'parameters'+str(i+1)+'.csv', sep = ';')
print "Posterior samples are saved to file."
if deviance == True:
for i in range(numberOfChains):
dev = models[i].mc.db.trace('deviance')()
numpy.savetxt('deviance'+str(i+1)+'.csv', dev, delimiter=";")
print "Deviance values are saved to file"
# if ss.is_group_model:
# print "DIC: %f" % ss.mc.dic