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
|
# This script is used to parse BOOST special function test data into something
# we can easily import in numpy. It is ugly as hell, but it works.
from __future__ import division, print_function, absolute_import
import re
import os
from numpy import arccosh, log1p, expm1, arcsinh, arctanh
from scipy.special import digamma, erfc, expi, erf, expn, gamma, gammaln, iv
# Where to put the data (directory will be created)
DATA_DIR = 'data'
# where to pull out boost data: assume a SVN checkout of boost (here in
# /Users/david/src/dev/boost/trunk)
BOOST_SRC = "boostmath/test"
CXX_COMMENT = re.compile(r'^\s+//')
DATA_REGEX = re.compile(r'^\s+/*\{*\s*SC_')
ITEM_REGEX = re.compile(r'[+-]?\d*\.?\d+(?:[eE][+-]?\d+)?')
HEADER_REGEX = re.compile(
r'const boost::array\<boost::array\<.*, (\d+)\>, (\d+)\> ([a-zA-Z_\d]+)')
# List of boost test data files to parse
DATA_FILES = [
'acosh_data.ipp',
'asinh_data.ipp',
'assoc_legendre_p.ipp',
'atanh_data.ipp',
'bessel_i_data.ipp',
'bessel_i_int_data.ipp',
'bessel_j_data.ipp',
'bessel_j_int_data.ipp',
'bessel_j_large_data.ipp',
'bessel_k_data.ipp',
'bessel_k_int_data.ipp',
'bessel_y01_data.ipp',
'bessel_yn_data.ipp',
'bessel_yv_data.ipp',
'beta_exp_data.ipp',
'beta_med_data.ipp',
'beta_small_data.ipp',
'binomial_data.ipp',
'binomial_large_data.ipp',
'binomial_quantile.ipp',
'cbrt_data.ipp',
'digamma_data.ipp',
'digamma_neg_data.ipp',
'digamma_root_data.ipp',
'digamma_small_data.ipp',
'ellint_e2_data.ipp',
'ellint_e_data.ipp',
'ellint_f_data.ipp',
'ellint_k_data.ipp',
'ellint_pi2_data.ipp',
'ellint_pi3_data.ipp',
'ellint_pi3_large_data.ipp',
'ellint_rc_data.ipp',
'ellint_rd_data.ipp',
'ellint_rf_data.ipp',
'ellint_rj_data.ipp',
'erfc_inv_big_data.ipp',
'erfc_inv_data.ipp',
'erf_data.ipp',
'erf_inv_data.ipp',
'erf_large_data.ipp',
'erf_small_data.ipp',
'expint_1_data.ipp',
'expint_data.ipp',
'expinti_data_double.ipp',
'expinti_data.ipp',
'expinti_data_long.ipp',
'expint_small_data.ipp',
'gamma_inv_big_data.ipp',
'gamma_inv_data.ipp',
'gamma_inv_small_data.ipp',
'hermite.ipp',
'hypergeometric_dist_data2.ipp',
'hypergeometric_test_data.ipp',
'ibeta_data.ipp',
'ibeta_int_data.ipp',
'ibeta_inva_data.ipp',
'ibeta_inv_data.ipp',
'ibeta_large_data.ipp',
'ibeta_small_data.ipp',
'igamma_big_data.ipp',
'igamma_int_data.ipp',
'igamma_inva_data.ipp',
'igamma_med_data.ipp',
'igamma_small_data.ipp',
'laguerre2.ipp',
'laguerre3.ipp',
'legendre_p.ipp',
'legendre_p_large.ipp',
'log1p_expm1_data.ipp',
'ncbeta_big.ipp',
'ncbeta.ipp',
'nccs_big.ipp',
'nccs.ipp',
'nct.ipp',
'negative_binomial_quantile.ipp',
'poisson_quantile.ipp',
'powm1_sqrtp1m1_test.cpp',
'sph_bessel_data.ipp',
'spherical_harmonic.ipp',
'sph_neumann_data.ipp',
#'test_bessel_i.cpp',
#'test_bessel_j.cpp',
#'test_bessel_k.cpp',
#'test_bessel_y.cpp',
# Those 3 files use arithmetic operations whithin the data, so we can't parse
# them naively
#'test_ellint_1.cpp',
#'test_ellint_2.cpp',
#'test_ellint_3.cpp',
'test_gamma_data.ipp',
'tgamma_delta_ratio_data.ipp',
'tgamma_delta_ratio_int2.ipp',
'tgamma_delta_ratio_int.ipp',
'tgamma_ratio_data.ipp',
'zeta_1_below_data.ipp',
'zeta_1_up_data.ipp',
'zeta_data.ipp',
'zeta_neg_data.ipp',
]
def _raw_data(line):
items = line.split(',')
l = []
for item in items:
m = ITEM_REGEX.search(item)
if m:
q = m.group(0)
l.append(q)
return l
def parse_ipp_file(filename):
print(filename)
a = open(filename, 'r')
lines = a.readlines()
data = {}
i = 0
while (i < len(lines)):
line = lines[i]
m = HEADER_REGEX.search(line)
if m:
d = int(m.group(1))
n = int(m.group(2))
print("d = {0}, n = {1}".format(d, n))
cdata = []
i += 1
line = lines[i]
# Skip comments
while CXX_COMMENT.match(line):
i += 1
line = lines[i]
while DATA_REGEX.match(line):
cdata.append(_raw_data(line))
i += 1
line = lines[i]
# Skip comments
while CXX_COMMENT.match(line):
i += 1
line = lines[i]
if not len(cdata) == n:
raise ValueError("parsed data: %d, expected %d" % (len(cdata), n))
data[m.group(3)] = cdata
else:
i += 1
return data
def dump_dataset(filename, data):
fid = open(filename, 'w')
try:
for line in data:
fid.write("%s\n" % " ".join(line))
finally:
fid.close()
def dump_datasets(filename):
base, ext = os.path.splitext(os.path.basename(filename))
base += '_%s' % ext[1:]
datadir = os.path.join(DATA_DIR, base)
os.makedirs(datadir)
datasets = parse_ipp_file(filename)
for k, d in datasets.items():
print(k, len(d))
dfilename = os.path.join(datadir, k) + '.txt'
dump_dataset(dfilename, d)
if __name__ == '__main__':
for filename in DATA_FILES:
filename = os.path.join(BOOST_SRC, filename)
print("================= %s ===============" % filename)
dump_datasets(filename)
|
