Codebase list cppad / upstream/2015.00.00.7 doc / mul_level_adolc.cpp.xml
upstream/2015.00.00.7

Tree @upstream/2015.00.00.7 (Download .tar.gz)

mul_level_adolc.cpp.xml @upstream/2015.00.00.7 — raw · history · blame 

  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
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
<?xml version='1.0'?>
<html xmlns='http://www.w3.org/1999/xhtml'
      xmlns:math='http://www.w3.org/1998/Math/MathML'
>
<head>
<title>Using Adolc with Multiple Levels of Taping: Example and Test</title>
<meta http-equiv='Content-Type' content='text/html' charset='utf-8'/>
<meta name="description" id="description" content="Using Adolc with Multiple Levels of Taping: Example and Test"/>
<meta name="keywords" id="keywords" content=" using adolc with multiple levels of taping: example and test Adolc level purpose memory management configuration requirement "/>
<style type='text/css'>
body { color : black }
body { background-color : white }
A:link { color : blue }
A:visited { color : purple }
A:active { color : purple }
</style>
<script type='text/javascript' language='JavaScript' src='_mul_level_adolc.cpp_xml.js'>
</script>
</head>
<body>
<table><tr>
<td>
<a href="http://www.coin-or.org/CppAD/" target="_top"><img border="0" src="_image.gif"/></a>
</td>
<td><a href="base_adolc.hpp.xml" target="_top">Prev</a>
</td><td><a href="base_float.hpp.xml" target="_top">Next</a>
</td><td>
<select onchange='choose_across0(this)'>
<option>Index-&gt;</option>
<option>contents</option>
<option>reference</option>
<option>index</option>
<option>search</option>
<option>external</option>
</select>
</td>
<td>
<select onchange='choose_up0(this)'>
<option>Up-&gt;</option>
<option>CppAD</option>
<option>AD</option>
<option>base_require</option>
<option>base_example</option>
<option>base_adolc.hpp</option>
<option>mul_level_adolc.cpp</option>
</select>
</td>
<td>
<select onchange='choose_down3(this)'>
<option>base_require-&gt;</option>
<option>base_member</option>
<option>base_cond_exp</option>
<option>base_identical</option>
<option>base_ordered</option>
<option>base_std_math</option>
<option>base_example</option>
</select>
</td>
<td>
<select onchange='choose_down2(this)'>
<option>base_example-&gt;</option>
<option>base_alloc.hpp</option>
<option>base_require.cpp</option>
<option>base_adolc.hpp</option>
<option>base_float.hpp</option>
<option>base_double.hpp</option>
<option>base_complex.hpp</option>
</select>
</td>
<td>
<select onchange='choose_down1(this)'>
<option>base_adolc.hpp-&gt;</option>
<option>mul_level_adolc.cpp</option>
</select>
</td>
<td>mul_level_adolc.cpp</td>
<td>
<select onchange='choose_current0(this)'>
<option>Headings-&gt;</option>
<option>Purpose</option>
<option>Memory Management</option>
<option>Configuration Requirement</option>
</select>
</td>
</tr></table><br/>



<center><b><big><big>Using Adolc with Multiple Levels of Taping: Example and Test</big></big></b></center>
<br/>
<b><big><a name="Purpose" id="Purpose">Purpose</a></big></b>
<br/>
In this example, we use <code><font color="blue">AD&lt; adouble&gt; &gt;</font></code> (level two taping),
the compute values of the function 
<math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><mrow>
<mi mathvariant='italic'>f</mi>
<mo stretchy="false">:</mo>
<msup><mrow><mstyle mathvariant='bold'><mi mathvariant='bold'>R</mi>
</mstyle></mrow>
<mi mathvariant='italic'>n</mi>
</msup>
<mo stretchy="false">&#x02192;</mo>
<mrow><mstyle mathvariant='bold'><mi mathvariant='bold'>R</mi>
</mstyle></mrow>
</mrow></math>

 where 

<math xmlns="http://www.w3.org/1998/Math/MathML" display="block"><mrow>
<mi mathvariant='italic'>f</mi>
<mo stretchy="false">(</mo>
<mi mathvariant='italic'>x</mi>
<mo stretchy="false">)</mo>
<mo stretchy="false">=</mo>
<mfrac><mrow><mn>1</mn>
</mrow>
<mrow><mn>2</mn>
</mrow>
</mfrac>
<mrow><mo stretchy="true">(</mo><mrow><msubsup><mi mathvariant='italic'>x</mi>
<mn>0</mn>
<mn>2</mn>
</msubsup>
<mo stretchy="false">+</mo>
<mo stretchy="false">&#x022EF;</mo>
<mo stretchy="false">+</mo>
<msubsup><mi mathvariant='italic'>x</mi>
<mrow><mi mathvariant='italic'>n</mi>
<mn>-1</mn>
</mrow>
<mn>2</mn>
</msubsup>
</mrow><mo stretchy="true">)</mo></mrow>
</mrow></math>

We then use Adolc's <code><font color="blue">adouble</font></code> (level one taping) to compute
the directional derivative

<math xmlns="http://www.w3.org/1998/Math/MathML" display="block"><mrow>
<msup><mi mathvariant='italic'>f</mi>
<mrow><mo stretchy="false">(</mo>
<mn>1</mn>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo stretchy="false">(</mo>
<mi mathvariant='italic'>x</mi>
<mo stretchy="false">)</mo>
<mo stretchy="false">*</mo>
<mi mathvariant='italic'>v</mi>
<mo stretchy="false">=</mo>
<msub><mi mathvariant='italic'>x</mi>
<mn>0</mn>
</msub>
<msub><mi mathvariant='italic'>v</mi>
<mn>0</mn>
</msub>
<mo stretchy="false">+</mo>
<mo stretchy="false">&#x022EF;</mo>
<mo stretchy="false">+</mo>
<msub><mi mathvariant='italic'>x</mi>
<mrow><mi mathvariant='italic'>n</mi>
<mn>-1</mn>
</mrow>
</msub>
<msub><mi mathvariant='italic'>v</mi>
<mrow><mi mathvariant='italic'>n</mi>
<mn>-1</mn>
</mrow>
</msub>
</mrow></math>

.
where 
<math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><mrow>
<mi mathvariant='italic'>v</mi>
<mo stretchy="false">&#x02208;</mo>
<msup><mrow><mstyle mathvariant='bold'><mi mathvariant='bold'>R</mi>
</mstyle></mrow>
<mi mathvariant='italic'>n</mi>
</msup>
</mrow></math>

.
We then use <code><font color="blue">double</font></code> (no taping) to compute

<math xmlns="http://www.w3.org/1998/Math/MathML" display="block"><mrow>
<mfrac><mrow><mi mathvariant='italic'>d</mi>
</mrow>
<mrow><mi mathvariant='italic'>dx</mi>
</mrow>
</mfrac>
<mrow><mo stretchy="true">[</mo><mrow><msup><mi mathvariant='italic'>f</mi>
<mrow><mo stretchy="false">(</mo>
<mn>1</mn>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo stretchy="false">(</mo>
<mi mathvariant='italic'>x</mi>
<mo stretchy="false">)</mo>
<mo stretchy="false">*</mo>
<mi mathvariant='italic'>v</mi>
</mrow><mo stretchy="true">]</mo></mrow>
<mo stretchy="false">=</mo>
<mi mathvariant='italic'>v</mi>
</mrow></math>

This is only meant as an example of multiple levels of taping.
The example <a href="hes_times_dir.cpp.xml" target="_top"><span style='white-space: nowrap'>hes_times_dir.cpp</span></a>
 computes the same value more
efficiently by using the identity:

<math xmlns="http://www.w3.org/1998/Math/MathML" display="block"><mrow>
<mfrac><mrow><mi mathvariant='italic'>d</mi>
</mrow>
<mrow><mi mathvariant='italic'>dx</mi>
</mrow>
</mfrac>
<mrow><mo stretchy="true">[</mo><mrow><msup><mi mathvariant='italic'>f</mi>
<mrow><mo stretchy="false">(</mo>
<mn>1</mn>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo stretchy="false">(</mo>
<mi mathvariant='italic'>x</mi>
<mo stretchy="false">)</mo>
<mo stretchy="false">*</mo>
<mi mathvariant='italic'>v</mi>
</mrow><mo stretchy="true">]</mo></mrow>
<mo stretchy="false">=</mo>
<msup><mi mathvariant='italic'>f</mi>
<mrow><mo stretchy="false">(</mo>
<mn>2</mn>
<mo stretchy="false">)</mo>
</mrow>
</msup>
<mo stretchy="false">(</mo>
<mi mathvariant='italic'>x</mi>
<mo stretchy="false">)</mo>
<mo stretchy="false">*</mo>
<mi mathvariant='italic'>v</mi>
</mrow></math>

The example <a href="mul_level.cpp.xml" target="_top"><span style='white-space: nowrap'>mul_level.cpp</span></a>
 computes the same values using
<code><font color="blue">AD&lt; AD&lt;double&gt; &gt;</font></code> and <code><font color="blue">AD&lt;double&gt;</font></code>.

<br/>
<br/>
<b><big><a name="Memory Management" id="Memory Management">Memory Management</a></big></b>
<br/>
Adolc uses raw memory arrays that depend on the number of 
dependent and independent variables.
The memory management utility <a href="omp_alloc.xml" target="_top"><span style='white-space: nowrap'>omp_alloc</span></a>
 
is used to manage this memory allocation.

<br/>
<br/>
<b><big><a name="Configuration Requirement" id="Configuration Requirement">Configuration Requirement</a></big></b>
<br/>
This example will be compiled and tested provided that
the value <a href="ipopt_prefix.xml" target="_top"><span style='white-space: nowrap'>ipopt_prefix</span></a>
 is specified on the 
<a href="cmake.xml" target="_top"><span style='white-space: nowrap'>cmake</span></a>
 command line.

<code><font color="blue">
<pre style='display:inline'> 
# include &lt;adolc/adouble.h&gt;
# include &lt;adolc/taping.h&gt;
# include &lt;adolc/interfaces.h&gt;

// adouble definitions not in Adolc distribution and 
// required in order to use CppAD::<a href="ad.xml" target="_top">AD</a>&lt;adouble&gt;
# include &lt;cppad/example/base_adolc.hpp&gt;

# include &lt;cppad/cppad.hpp&gt;

namespace { 
	// f(x) = |x|^2 / 2 = .5 * ( x[0]^2 + ... + x[n-1]^2 )
	template &lt;class Type&gt;
	Type f(CPPAD_TESTVECTOR(Type) &amp;x)
	{	Type sum;

		sum  = 0.;
		size_t i = size_t(x.size());
		for(i = 0; i &lt; size_t(x.size()); i++)
			sum += x[i] * x[i];

		return .5 * sum;
	} 
}

bool mul_level_adolc(void) 
{	bool ok = true;                // initialize test result
	using CppAD::omp_alloc;        // The CppAD memory allocator

	typedef adouble             ADdouble;  // for first level of taping
	typedef CppAD::<a href="ad.xml" target="_top">AD</a>&lt;ADdouble&gt; ADDdouble; // for second level of taping
	size_t n = 5;                          // number independent variables
	size_t j;

	<a href="testvector.xml" target="_top">CPPAD_TESTVECTOR</a>(double)       x(n);
	<a href="testvector.xml" target="_top">CPPAD_TESTVECTOR</a>(ADdouble)   a_x(n);
	<a href="testvector.xml" target="_top">CPPAD_TESTVECTOR</a>(ADDdouble) aa_x(n);

	// Values for the independent variables while taping the function f(x)
	for(j = 0; j &lt; n; j++)
		aa_x[j] = double(j);
	// Declare the independent variable for taping f(x)
	CppAD::<a href="independent.xml" target="_top">Independent</a>(aa_x); 

	// Use <a href="ad.xml" target="_top">AD</a>&lt;adouble&gt; to tape the evaluation of f(x)
	<a href="testvector.xml" target="_top">CPPAD_TESTVECTOR</a>(ADDdouble) aa_f(1); 
	aa_f[0] = f(aa_x); 

	// Declare a_F as the corresponding <a href="funconstruct.xml" target="_top">ADFun</a>&lt;adouble&gt; function f(x)
	// (make sure we do not run zero order forward during constructor)
	CppAD::<a href="funconstruct.xml" target="_top">ADFun</a>&lt;ADdouble&gt; a_F;
	a_F.Dependent(aa_x, aa_f);

	// Value of the independent variables whitle taping f'(x) * v
	int tag = 0; 
	int keep = 1;
	trace_on(tag, keep);
	for(j = 0; j &lt; n; j++)
		a_x[j] &lt;&lt;= double(j);

	// set the argument value x for computing f'(x) * v
	a_F.<a href="forward.xml" target="_top">Forward</a>(0, a_x);

	// compute f'(x) * v
	<a href="testvector.xml" target="_top">CPPAD_TESTVECTOR</a>(ADdouble) a_v(n);
	<a href="testvector.xml" target="_top">CPPAD_TESTVECTOR</a>(ADdouble) a_df(1);
	for(j = 0; j &lt; n; j++)
		a_v[j] = double(n - j);
	a_df = a_F.<a href="forward.xml" target="_top">Forward</a>(1, a_v); 

	// declare Adolc function corresponding to f'(x) * v
	double df; 
	a_df[0] &gt;&gt;= df;
	trace_off();

	// compute the d/dx of f'(x) * v = f''(x) * v
	size_t m      = 1;                     // # dependent in f'(x) * v

	// w = new double[capacity] where capacity &gt;= m
	size_t capacity;
	double* w  = omp_alloc::create_array&lt;double&gt;(m, capacity);
	// w = new double[capacity] where capacity &gt;= n
	double* dw = omp_alloc::create_array&lt;double&gt;(n, capacity);

	w[0]  = 1.;
	fos_reverse(tag, int(m), int(n), w, dw);

	for(j = 0; j &lt; n; j++)
	{	double vj = a_v[j].value();
		ok &amp;= CppAD::<a href="nearequal.xml" target="_top">NearEqual</a>(dw[j], vj, 1e-10, 1e-10);
	}

	// make memory avaialble for other use by this thread
	omp_alloc::delete_array(w);
	omp_alloc::delete_array(dw);
	return ok;
}
</pre>

</font></code>


<hr/>Input File: example/mul_level_adolc.cpp

</body>
</html>