-
Notifications
You must be signed in to change notification settings - Fork 1
/
float_vectors.py
229 lines (178 loc) · 7.14 KB
/
float_vectors.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
#!/usr/bin/env python3
##############################################################################
## ##
## FP_TEST_GENERATOR ##
## ##
## Copyright (C) 2020, Florian Schanda ##
## ##
## This file is part of FP_Test_Generator. ##
## ##
## FP_Test_Generator is free software: you can redistribute it and/or ##
## modify it under the terms of the GNU General Public License as ##
## published by the Free Software Foundation, either version 3 of the ##
## License, or (at your option) any later version. ##
## ##
## FP_Test_Generator is distributed in the hope that it will be useful, ##
## but WITHOUT ANY WARRANTY; without even the implied warranty of ##
## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the ##
## GNU General Public License for more details. ##
## ##
## You should have received a copy of the GNU General Public License ##
## along with FP_Test_Generator. If not, see ##
## <http://www.gnu.org/licenses/>. ##
## ##
##############################################################################
from functools import partial
from mpf.floats import MPF, RM_RNE, fp_nextUp, fp_nextDown
from mpf.rationals import Rational
from core import Vector
def build_zero(eb, sb, _, sign):
rv = MPF(eb, sb)
rv.set_zero(sign)
return rv
def build_min_subnormal(eb, sb, _, sign):
rv = MPF(eb, sb)
rv.pack(sign, 0, 1)
return rv
def build_rnd_subnormal(eb, sb, rng, sign):
rv = MPF(eb, sb)
rv.pack(sign, 0, rng.random_int(2, 2 ** rv.t - 2))
return rv
def build_max_subnormal(eb, sb, _, sign):
rv = MPF(eb, sb)
rv.pack(sign, 0, 2 ** rv.t - 1)
return rv
def build_min_normal(eb, sb, _, sign):
rv = MPF(eb, sb)
rv.pack(sign, 1, 0)
return rv
def build_rnd_normal_small(eb, sb, rng, sign):
rv = MPF(eb, sb)
rv.pack(sign,
rng.random_int(1, 2 ** (rv.w - 1)),
rng.random_int(0, 2 ** rv.t - 1))
return rv
def build_rnd_normal_large(eb, sb, rng, sign):
rv = MPF(eb, sb)
rv.pack(sign,
rng.random_int(2 ** (rv.w - 1), 2 ** rv.w - 2),
rng.random_int(0, 2 ** rv.t - 1))
return rv
def build_one(eb, sb, _, sign):
rv = MPF(eb, sb)
rv.from_rational(RM_RNE, Rational(1))
rv.set_sign_bit(sign)
return rv
def build_one_succ(eb, sb, rng, sign):
rv = build_one(eb, sb, rng, sign)
return fp_nextUp(rv)
def build_one_pred(eb, sb, rng, sign):
rv = build_one(eb, sb, rng, sign)
return fp_nextDown(rv)
def build_max_normal(eb, sb, _, sign):
rv = MPF(eb, sb)
rv.pack(sign, 2 ** rv.w - 2, 2 ** rv.t - 1)
return rv
def build_inf(eb, sb, _, sign):
rv = MPF(eb, sb)
rv.pack(sign, 2 ** rv.w - 1, 0)
return rv
def build_nan(eb, sb, rng):
rv = MPF(eb, sb)
rv.pack(rng.random_int(0, 1),
2 ** rv.w - 1,
rng.random_int(1, 2 ** rv.t - 1))
return rv
fp_test_points = {
# Zeros
"+0" : partial(build_zero, sign=0),
"-0" : partial(build_zero, sign=1),
# Subnormals
"+min_subnormal" : partial(build_min_subnormal, sign=0),
"+rnd_subnormal" : partial(build_rnd_subnormal, sign=0),
"+max_subnormal" : partial(build_max_subnormal, sign=0),
"-min_subnormal" : partial(build_min_subnormal, sign=1),
"-rnd_subnormal" : partial(build_rnd_subnormal, sign=1),
"-max_subnormal" : partial(build_max_subnormal, sign=1),
# Normals
"+min_normal" : partial(build_min_normal, sign=0),
"+rnd_normal_small" : partial(build_rnd_normal_small, sign=0),
"+rnd_normal_large" : partial(build_rnd_normal_large, sign=0),
"+max_normal" : partial(build_max_normal, sign=0),
"+1" : partial(build_one, sign=0),
"nextup(+1)" : partial(build_one_succ, sign=0),
"nextdown(+1)" : partial(build_one_pred, sign=0),
"-min_normal" : partial(build_min_normal, sign=1),
"-rnd_normal_small" : partial(build_rnd_normal_small, sign=1),
"-rnd_normal_large" : partial(build_rnd_normal_large, sign=1),
"-max_normal" : partial(build_max_normal, sign=1),
"-1" : partial(build_one, sign=1),
"nextup(-1)" : partial(build_one_succ, sign=1),
"nextdown(-1)" : partial(build_one_pred, sign=1),
# Infinities
"+inf" : partial(build_inf, sign=0),
"-inf" : partial(build_inf, sign=1),
# NaN
"NaN" : build_nan,
}
reduced_set = frozenset(["+0", "-0",
"+rnd_subnormal", "-rnd_subnormal",
"+rnd_normal_small", "-rnd_normal_small",
"+inf", "-inf",
"NaN"])
class Float_Vector(Vector):
def __init__(self):
self.vec = []
def add_item(self, kind):
assert kind in fp_test_points
self.vec.append(kind)
def __str__(self):
return "Float_Vector<%s>" % ", ".join(self.vec)
@classmethod
def generate(cls, eb, sb, size, reduced):
assert isinstance(eb, int)
assert isinstance(sb, int)
assert isinstance(size, int)
assert isinstance(reduced, bool)
assert size >= 1
if reduced:
all_kinds = sorted(reduced_set)
else:
all_kinds = list(sorted(fp_test_points))
kinds = [0] * size
def increment(p):
if p == size:
return False
kinds[p] += 1
if kinds[p] == len(all_kinds):
kinds[p] = 0
return increment(p + 1)
else:
return True
while True:
vec = Float_Vector()
for k in kinds:
vec.add_item(all_kinds[k])
yield vec
if not increment(0):
return
class Float_Vector_With_RM(Float_Vector):
def __init__(self, rm):
super().__init__()
assert rm in MPF.ROUNDING_MODES
self.rm = rm
def __str__(self):
return "Float_Vector_With_RM<%s,%s>" % (self.rm,
", ".join(self.vec))
@classmethod
def generate(cls, eb, sb, size, reduced):
assert isinstance(eb, int)
assert isinstance(sb, int)
assert isinstance(size, int)
assert isinstance(reduced, bool)
assert size >= 1
for fv in Float_Vector.generate(eb, sb, size, reduced):
for rm in MPF.ROUNDING_MODES:
vec = Float_Vector_With_RM(rm)
vec.vec = fv.vec
yield vec