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
|
<html>
<head>
<title>Blitz++ User's Guide </title>
</head>
<body fgcolor="#27408B" bgcolor="#FFFAF0" >
<hr>
<ul>
<li> <a href="blitz11.html">Next chapter</a>
<li> <a href="blitz09.html">Previous chapter</a>
<li> <a href="blitz.html">Table of contents</a>
</ul>
<hr>
<a name="l135"></a>
<h1>Chapter 10: Numeric properties</h1>
<a name="numinquire"></a>
<br><br><br><table width="100%" border="0" cellpadding=10 align=center><tr><td align="left" bgcolor="#0b6698"><font color="#ffffff" face="Helvetica" size=+5>10.1: Introduction</font></td></tr></table><br><a name="l136"></a>
<p><!-- BZINDEX numeric limits --><a name="index00460">
<p>Blitz++ provides a set of functions to
access numeric properties of intrinsic types.
They are provided as an alternative to the somewhat
klunky <code>numeric_limits<T>::yadda_yadda</code> syntax
provided by the ISO/ANSI C++ standard. Where a
similar Fortran 90 function exists, the same name
has been used.
<p>The argument in all cases is a dummy of the appropriate
type.
<p>All functions described in this section assume that
<code>numeric_limits<T></code> has been specialized for the
appropriate case. If not, the results are not useful.
The standard requires that <code>numeric_limits<T></code> be
specialized for all the intrinsic numeric types
(float, double, int, bool, unsigned int, etc.).
<p><!-- BZINDEX numinquire.h --><a name="index00461">
<p>To use these functions, you must first include the
header <code><blitz/numinquire.h></code>. Also, note that
these functions may be unavailable if your compiler
is non-ANSI compliant. If the preprocessor symbol
<code>BZ_HAVE_NUMERIC_LIMITS</code> is false, then these
functions are unavailable.
<p>
<p><br><br><br><table width="100%" border="0" cellpadding=10 align=center><tr><td align="left" bgcolor="#0b6698"><font color="#ffffff" face="Helvetica" size=+5>10.2: Function descriptions</font></td></tr></table><br><a name="l137"></a>
<p><!-- BZINDEX denorm_min() --><a name="index00462">
<strong><pre>T denorm_min(T) throw;
</pre></strong>
Minimum positive denormalized value. Available for floating-point
types only.
<p><!-- BZINDEX digits() --><a name="index00463">
<strong><pre>int digits(T);
</pre></strong>
The number of radix digits (read: bits) in the mantissa. Also
works for integer types. The official definition is "number of
radix digits that can be represented without change".
<p><!-- BZINDEX digits10() --><a name="index00464">
<strong><pre>int digits10(T);
</pre></strong>
The number of base-10 digits that can be represented without
change.
<p><!-- BZINDEX epsilon() --><a name="index00465">
<strong><pre>T epsilon(T);
</pre></strong>
The smallest amount which can be added to 1 to produce
a result which is not 1. Floating-point types only.
<p><!-- BZINDEX denormalized values --><a name="index00466">
<!-- BZINDEX has_denorm() --><a name="index00467">
<strong><pre>bool has_denorm(T);
</pre></strong>
True if the representation allows denormalized values (floating-point
only).
<p><!-- BZINDEX denormalization loss --><a name="index00468">
<!-- BZINDEX has_denorm_loss() --><a name="index00469">
<strong><pre>bool has_denorm_loss(T);
</pre></strong>
True if a loss of precision is detected as a denormalization loss, rather
than as an inexact result (floating-point only).
<p><!-- BZINDEX infinity -- has_infinity() --><a name="index00470">
<!-- BZINDEX has_infinity() --><a name="index00471">
<strong><pre>bool has_infinity(T);
</pre></strong>
True if there is a special representation for the value "infinity".
If true, the representation can be obtained by calling <code>infinity(T)</code>.
<p><!-- BZINDEX has_quiet_NaN() --><a name="index00472">
<!-- BZINDEX NaN -- has_quiet_NaN() --><a name="index00473">
<strong><pre>bool has_quiet_NaN(T);
</pre></strong>
True if there is a special representation for a quiet (non-signalling)
Not A Number (NaN). If so, use the function <code>quiet_NaN(T)</code> to obtain
it.
<p><!-- BZINDEX NaN -- has_signalling_NaN() --><a name="index00474">
<!-- BZINDEX has_signalling_NaN() --><a name="index00475">
<strong><pre>bool has_signaling_NaN(T);
bool has_signalling_NaN(T);
</pre></strong>
True if there is a special representation for a signalling
Not A Number (NaN). If so, use the function <code>signalling_NaN(T)</code>
to obtain it.
<p><!-- BZINDEX huge() --><a name="index00476">
<!-- BZINDEX maximum value of a type --><a name="index00477">
<strong><pre>T huge(T) throw;
</pre></strong>
Returns the maximum finite representable value. Equivalent to
<code>CHAR_MAX</code>, <code>SHRT_MAX</code>, <code>FLT_MAX</code>, etc. For floating types
with denormalization, the maximum positive <strong>normalized</strong> value is
returned.
<p><!-- BZINDEX infinity() --><a name="index00478">
<strong><pre>T infinity(T) throw;
</pre></strong>
Returns the representation of positive infinity, if available.
Note that you should check availability with <code>has_infinity(T)</code>
before calling this function.
<p><!-- BZINDEX is_bounded() --><a name="index00479">
<strong><pre>bool is_bounded(T);
</pre></strong>
True if the set of values represented by the type is finite.
All built-in types are bounded. (This function was provided
so that e.g. arbitrary precision types could be distinguished).
<p><!-- BZINDEX is_exact() --><a name="index00480">
<strong><pre>bool is_exact(T);
</pre></strong>
True if the representation is exact. All integer types are
exact; floating-point types generally aren't. A rational
arithmetic type could be exact.
<p><!-- BZINDEX IEC 559 --><a name="index00481">
<!-- BZINDEX is_iec559() --><a name="index00482">
<strong><pre>bool is_iec559(T);
</pre></strong>
True if the type conforms to the IEC 559 standard. IEC is
the International Electrotechnical Commission. Note that
IEC 559 is the same as IEEE 754. Only relevant for floating types.
<p><!-- BZINDEX is_integer() --><a name="index00483">
<strong><pre>bool is_integer(T);
</pre></strong>
True if the type is integer.
<p><!-- BZINDEX is_modulo() --><a name="index00484">
<strong><pre>bool is_modulo(T);
</pre></strong>
True if the type is modulo. Integer types are usually
modulo: if you add two integers, they might wrap around
and give you a small result. (Some special kinds of
integers don't wrap around, but stop at an upper or
lower bound; this is called saturating arithmetic).
This is false for floating types.
<p><!-- BZINDEX is_signed() --><a name="index00485">
<!-- BZINDEX signed -- is_signed() --><a name="index00486">
<strong><pre>bool is_signed(T);
</pre></strong>
True if the type is signed (i.e. can handle both positive
and negative values).
<p><strong><pre>int max_exponent(T);
</pre></strong>
The maximum exponent (Max_exp) is the maximum positive integer such that
the radix (read: 2) raised to the power Max_exp-1 is a representable,
finite floating point number. Floating types only. <!-- BZINDEX max_exponent() --><a name="index00487">
<p><!-- BZINDEX max_exponent10() --><a name="index00488">
<strong><pre>int max_exponent10(T);
</pre></strong>
The maximum base-10 exponent (Max_exp10) is the maximum positive integer
such that 10 raised to the power Max_exp10 is a representable,
finite floating point number. Floating types only.
<p><!-- BZINDEX min_exponent() --><a name="index00489">
<strong><pre>int min_exponent(T);
</pre></strong>
The minimum exponent (Min_exp) is the minimum negative integer
such that the radix (read: 2) raised to the power Min_exp-1 is
a <strong>normalized</strong> floating point number. Floating types only.
<p><!-- BZINDEX min_exponent10() --><a name="index00490">
<strong><pre>int min_exponent10(T);
</pre></strong>
The minimum base-10 exponent (Min_exp10) is the minimum negative
integer such that 10 raised to the power Min_exp10 is in the
range of <strong>normalized</strong> floating point numbers.
<p><!-- BZINDEX neghuge() --><a name="index00491">
<!-- BZINDEX maximally negative value -- neghuge() --><a name="index00492">
<strong><pre>T neghuge(T);
</pre></strong>
This returns the maximally negative value for a type.
For integers, this is the same as min(). For floating-point
types, it is <code>- huge(T())</code>.
<p><!-- BZINDEX one() --><a name="index00493">
<strong><pre>T one(T);
</pre></strong>
Returns a representation for "1".
<p><!-- BZINDEX precision() --><a name="index00494">
<strong><pre>int precision(T);
</pre></strong>
Same as <code>digits10()</code>.
<p><!-- BZINDEX quiet_NaN() --><a name="index00495">
<!-- BZINDEX NaN -- quiet_NaN() --><a name="index00496">
<strong><pre>T quiet_NaN(T) throw;
</pre></strong>
Returns the representation for a quiet (non-signalling) Not A Number
(NaN), if available. You should check availability using the
<code>has_quiet_NaN(T)</code> function first.
<p><!-- BZINDEX radix() --><a name="index00497">
<strong><pre>int radix(T);
</pre></strong>
For floating-point types, this returns the radix (base) of the
exponent. For integers, it specifies the base of the representation.
<p><!-- BZINDEX range() --><a name="index00498">
<strong><pre>Range range(T);
</pre></strong>
Returns <code>Range(min_exponent10(T()), max_exponent10(T()))</code>, i.e.
the range of representable base-10 exponents.
<p><!-- BZINDEX round_error() --><a name="index00499">
<strong><pre>T round_error(T) throw;
</pre></strong>
Returns a measure of the maximum rounding error for floating-point types.
This will typically be <code>0.5</code>.
<p><!-- BZINDEX round_style() --><a name="index00500">
<strong><pre>std::float_round_style round_style(T);
</pre></strong>
Returns the current rounding style for floating-point arithmetic.
The possibilities are: <code>round_indeterminate</code> (i.e. don't have a clue),
<code>round_toward_zero</code>, <code>round_to_nearest</code> (round to nearest
representable value), <code>round_toward_infinity</code> (round toward
positive infinity), and <code>round_neg_infinity</code> (round toward
negative infinity).
<p><!-- BZINDEX signalling_NaN() --><a name="index00501">
<strong><pre>T signaling_NaN(T) throw;
T signalling_NaN(T) throw;
</pre></strong>
Returns the representation for a signalling Not A Number (NaN),
if available. You should check availability by calling
<pre>has_signalling_NaN(T)</pre> first.
<p><!-- BZINDEX tiny() --><a name="index00502">
<!-- BZINDEX minimum finite value -- tiny() --><a name="index00503">
<strong><pre>T tiny(T);
</pre></strong>
For integer types, this returns the minimum finite value, which may be
negative. For floating types, it returns the minimum positive value.
For floating types with denormalization, the function
returns the minimum positive <strong>normalized</strong> value.
<p><!-- BZINDEX tinyness_before() --><a name="index00504">
<strong><pre>T tinyness_before(T);
</pre></strong>
True if tinyness is detected before rounding. Other than this
description, I don't have a clue what this means; anyone have
a copy of IEC 559/IEEE 754 floating around?
<p><!-- BZINDEX traps() --><a name="index00505">
<strong><pre>T traps(T);
</pre></strong>
True if trapping is implemented for this type.
<p><!-- BZINDEX zero() --><a name="index00506">
<strong><pre>T zero(T);
</pre></strong>
Returns a representation for zero.
<p>
<p>
<hr>
<ul>
<li> <a href="blitz11.html">Next chapter</a>
<li> <a href="blitz09.html">Previous chapter</a>
<li> <a href="blitz.html">Table of contents</a>
</ul>
<hr>
</body>
</html>
|
