File: exam_pseries.cpp

package info (click to toggle)
ginac 1.8.9-1
  • links: PTS
  • area: main
  • in suites: forky, sid
  • size: 6,640 kB
  • sloc: cpp: 49,195; sh: 5,402; makefile: 448; python: 193
file content (439 lines) | stat: -rw-r--r-- 12,940 bytes parent folder | download | duplicates (2)
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
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
/** @File exam_pseries.cpp
 *
 *  Series expansion test (Laurent and Taylor series). */

/*
 *  GiNaC Copyright (C) 1999-2025 Johannes Gutenberg University Mainz, Germany
 *
 *  This program 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 2 of the License, or
 *  (at your option) any later version.
 *
 *  This program 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 this program; if not, write to the Free Software
 *  Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301  USA
 */

#include "ginac.h"
using namespace GiNaC;

#include <iostream>
using namespace std;

static symbol x("x");

static unsigned check_series(const ex &e, const ex &point, const ex &d, int order = 8)
{
	ex es = e.series(x==point, order);
	ex ep = ex_to<pseries>(es).convert_to_poly();
	if (!(ep - d).expand().is_zero()) {
		clog << "series expansion of " << e << " at " << point
		     << " erroneously returned " << ep << " (instead of " << d
		     << ")" << endl;
		clog << tree << (ep-d) << dflt;
		return 1;
	}
	return 0;
}

// Series expansion
static unsigned exam_series1()
{
	using GiNaC::log;

	symbol a("a");
	symbol b("b");
	unsigned result = 0;
	ex e, d;
	
	e = pow(a+b, x);
	d = 1 + Order(pow(x, 1));
	result += check_series(e, 0, d, 1);

	e = sin(x);
	d = x - pow(x, 3) / 6 + pow(x, 5) / 120 - pow(x, 7) / 5040 + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = cos(x);
	d = 1 - pow(x, 2) / 2 + pow(x, 4) / 24 - pow(x, 6) / 720 + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = exp(x);
	d = 1 + x + pow(x, 2) / 2 + pow(x, 3) / 6 + pow(x, 4) / 24 + pow(x, 5) / 120 + pow(x, 6) / 720 + pow(x, 7) / 5040 + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = pow(1 - x, -1);
	d = 1 + x + pow(x, 2) + pow(x, 3) + pow(x, 4) + pow(x, 5) + pow(x, 6) + pow(x, 7) + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = x + pow(x, -1);
	d = x + pow(x, -1);
	result += check_series(e, 0, d);
	
	e = x + pow(x, -1);
	d = 2 + pow(x-1, 2) - pow(x-1, 3) + pow(x-1, 4) - pow(x-1, 5) + pow(x-1, 6) - pow(x-1, 7) + Order(pow(x-1, 8));
	result += check_series(e, 1, d);
	
	e = pow(x + pow(x, 3), -1);
	d = pow(x, -1) - x + pow(x, 3) - pow(x, 5) + pow(x, 7) + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = pow(pow(x, 2) + pow(x, 4), -1);
	d = pow(x, -2) - 1 + pow(x, 2) - pow(x, 4) + pow(x, 6) + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = pow(sin(x), -2);
	d = pow(x, -2) + numeric(1,3) + pow(x, 2) / 15 + pow(x, 4) * 2/189 + pow(x, 6) / 675  + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = sin(x) / cos(x);
	d = x + pow(x, 3) / 3 + pow(x, 5) * 2/15 + pow(x, 7) * 17/315 + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = cos(x) / sin(x);
	d = pow(x, -1) - x / 3 - pow(x, 3) / 45 - pow(x, 5) * 2/945 - pow(x, 7) / 4725 + Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	e = pow(numeric(2), x);
	ex t = log(2) * x;
	d = 1 + t + pow(t, 2) / 2 + pow(t, 3) / 6 + pow(t, 4) / 24 + pow(t, 5) / 120 + pow(t, 6) / 720 + pow(t, 7) / 5040 + Order(pow(x, 8));
	result += check_series(e, 0, d.expand());
	
	e = pow(Pi, x);
	t = log(Pi) * x;
	d = 1 + t + pow(t, 2) / 2 + pow(t, 3) / 6 + pow(t, 4) / 24 + pow(t, 5) / 120 + pow(t, 6) / 720 + pow(t, 7) / 5040 + Order(pow(x, 8));
	result += check_series(e, 0, d.expand());
	
	e = log(x);
	d = e;
	result += check_series(e, 0, d, 1);
	result += check_series(e, 0, d, 2);
	
	e = pow(x, 8) * pow(pow(x,3)+ pow(x + pow(x,3), 2), -2);
	d = pow(x, 4) - 2*pow(x, 5) + Order(pow(x, 6));
	result += check_series(e, 0, d, 6);
	
	e = cos(x) * pow(sin(x)*(pow(x, 5) + 4 * pow(x, 2)), -3);
	d = pow(x, -9) / 64 - 3 * pow(x, -6) / 256 - pow(x, -5) / 960 + 535 * pow(x, -3) / 96768
	    + pow(x, -2) / 1280 - pow(x, -1) / 14400 - numeric(283, 129024) - 2143 * x / 5322240
	    + Order(pow(x, 2));
	result += check_series(e, 0, d, 2);
	
	e = sqrt(1+x*x) * sqrt(1+2*x*x);
	d = 1 + Order(pow(x, 2));
	result += check_series(e, 0, d, 2);

	e = pow(x, 4) * sin(a) + pow(x, 2);
	d = pow(x, 2) + Order(pow(x, 3));
	result += check_series(e, 0, d, 3);

	e = log(a*x + b*x*x*log(x));
	d = log(a*x) + b/a*log(x)*x - pow(b/a, 2)/2*pow(log(x)*x, 2) + Order(pow(x, 3));
	result += check_series(e, 0, d, 3);

	e = pow((x+a), b);
	d = pow(a, b) + (pow(a, b)*b/a)*x + (pow(a, b)*b*b/a/a/2 - pow(a, b)*b/a/a/2)*pow(x, 2) + Order(pow(x, 3));
	result += check_series(e, 0, d, 3);

	e = a * (1 / (x * sin(x)) - sin(x) / x);
	d = a * pow(x, -2) + Order(pow(x, -1));
	result += check_series(e, 0, d, -1);

	return result;
}

// Series addition
static unsigned exam_series2()
{
	unsigned result = 0;
	ex e, d;
	
	e = pow(sin(x), -1).series(x==0, 8) + pow(sin(-x), -1).series(x==0, 12);
	d = Order(pow(x, 8));
	result += check_series(e, 0, d);
	
	return result;
}

// Series multiplication
static unsigned exam_series3()
{
	unsigned result = 0;
	ex e, d;
	
	e = sin(x).series(x==0, 8) * pow(sin(x), -1).series(x==0, 12);
	d = 1 + Order(pow(x, 7));
	result += check_series(e, 0, d);
	
	return result;
}

// Series exponentiation
static unsigned exam_series4()
{
	using GiNaC::tgamma;
	unsigned result = 0;
	ex e, d;
	
	e = pow((2*cos(x)).series(x==0, 5), 2).series(x==0, 5);
	d = 4 - 4*pow(x, 2) + 4*pow(x, 4)/3 + Order(pow(x, 5));
	result += check_series(e, 0, d);
	
	e = pow(tgamma(x), 2).series(x==0, 2);
	d = pow(x,-2) - 2*Euler/x + (pow(Pi,2)/6+2*pow(Euler,2)) 
		+ x*(-4*pow(Euler, 3)/3 -pow(Pi,2)*Euler/3 - 2*zeta(3)/3) + Order(pow(x, 2));
	result += check_series(e, 0, d);
	
	return result;
}

// Order term handling
static unsigned exam_series5()
{
	unsigned result = 0;
	ex e, d;

	e = 1 + x + pow(x, 2) + pow(x, 3);
	d = Order(1);
	result += check_series(e, 0, d, 0);
	d = 1 + Order(x);
	result += check_series(e, 0, d, 1);
	d = 1 + x + Order(pow(x, 2));
	result += check_series(e, 0, d, 2);
	d = 1 + x + pow(x, 2) + Order(pow(x, 3));
	result += check_series(e, 0, d, 3);
	d = 1 + x + pow(x, 2) + pow(x, 3);
	result += check_series(e, 0, d, 4);
	return result;
}

// Series expansion of tgamma(-1)
static unsigned exam_series6()
{
	using GiNaC::tgamma;
	ex e = tgamma(2*x);
	ex d = pow(x+1,-1)*numeric(1,4) +
	       pow(x+1,0)*(numeric(3,4) -
	                   numeric(1,2)*Euler) +
	       pow(x+1,1)*(numeric(7,4) -
	                   numeric(3,2)*Euler +
	                   numeric(1,2)*pow(Euler,2) +
	                   numeric(1,12)*pow(Pi,2)) +
	       pow(x+1,2)*(numeric(15,4) -
	                   numeric(7,2)*Euler -
	                   numeric(1,3)*pow(Euler,3) +
	                   numeric(1,4)*pow(Pi,2) +
	                   numeric(3,2)*pow(Euler,2) -
	                   numeric(1,6)*pow(Pi,2)*Euler -
	                   numeric(2,3)*zeta(3)) +
	       pow(x+1,3)*(numeric(31,4) - pow(Euler,3) -
	                   numeric(15,2)*Euler +
	                   numeric(1,6)*pow(Euler,4) +
	                   numeric(7,2)*pow(Euler,2) +
	                   numeric(7,12)*pow(Pi,2) -
	                   numeric(1,2)*pow(Pi,2)*Euler -
	                   numeric(2)*zeta(3) +
	                   numeric(1,6)*pow(Euler,2)*pow(Pi,2) +
	                   numeric(1,40)*pow(Pi,4) +
	                   numeric(4,3)*zeta(3)*Euler) +
	       Order(pow(x+1,4));
	return check_series(e, -1, d, 4);
}
	
// Series expansion of tan(x==Pi/2)
static unsigned exam_series7()
{
	ex e = tan(x*Pi/2);
	ex d = pow(x-1,-1)/Pi*(-2) + pow(x-1,1)*Pi/6 + pow(x-1,3)*pow(Pi,3)/360
	      +pow(x-1,5)*pow(Pi,5)/15120 + pow(x-1,7)*pow(Pi,7)/604800
	      +Order(pow(x-1,9));
	return check_series(e,1,d,9);
}

// Series expansion of log(sin(x==0))
static unsigned exam_series8()
{
	ex e = log(sin(x));
	ex d = log(x) - pow(x,2)/6 - pow(x,4)/180 - pow(x,6)/2835 - pow(x,8)/37800 + Order(pow(x,9));
	return check_series(e,0,d,9);
}

// Series expansion of Li2(sin(x==0))
static unsigned exam_series9()
{
	ex e = Li2(sin(x));
	ex d = x + pow(x,2)/4 - pow(x,3)/18 - pow(x,4)/48
	       - 13*pow(x,5)/1800 - pow(x,6)/360 - 23*pow(x,7)/21168
	       + Order(pow(x,8));
	return check_series(e,0,d,8);
}

// Series expansion of Li2((x==2)^2), caring about branch-cut
static unsigned exam_series10()
{
	using GiNaC::log;

	ex e = Li2(pow(x,2));
	ex d = Li2(4) + (-log(3) + I*Pi*csgn(I-I*pow(x,2))) * (x-2)
	       + (numeric(-2,3) + log(3)/4 - I*Pi/4*csgn(I-I*pow(x,2))) * pow(x-2,2)
	       + (numeric(11,27) - log(3)/12 + I*Pi/12*csgn(I-I*pow(x,2))) * pow(x-2,3)
	       + (numeric(-155,648) + log(3)/32 - I*Pi/32*csgn(I-I*pow(x,2))) * pow(x-2,4)
	       + Order(pow(x-2,5));
	return check_series(e,2,d,5);
}

// Series expansion of logarithms around branch points
static unsigned exam_series11()
{
	using GiNaC::log;

	unsigned result = 0;
	ex e, d;
	symbol a("a");
	
	e = log(x);
	d = log(x);
	result += check_series(e,0,d,5);
	
	e = log(3/x);
	d = log(3)-log(x);
	result += check_series(e,0,d,5);
	
	e = log(3*pow(x,2));
	d = log(3)+2*log(x);
	result += check_series(e,0,d,5);
	
	// These ones must not be expanded because it would result in a branch cut
	// running in the wrong direction. (Other systems tend to get this wrong.)
	e = log(-x);
	d = e;
	result += check_series(e,0,d,5);
	
	e = log(I*(x-123));
	d = e;
	result += check_series(e,123,d,5);
	
	e = log(a*x);
	d = e;  // we don't know anything about a!
	result += check_series(e,0,d,5);
	
	e = log((1-x)/x);
	d = log(1-x) - (x-1) + pow(x-1,2)/2 - pow(x-1,3)/3  + pow(x-1,4)/4 + Order(pow(x-1,5));
	result += check_series(e,1,d,5);
	
	return result;
}

// Series expansion of other functions around branch points
static unsigned exam_series12()
{
	using GiNaC::log;
	using GiNaC::atanh;

	unsigned result = 0;
	ex e, d;
	
	// NB: Mma and Maple give different results, but they agree if one
	// takes into account that by assumption |x|<1.
	e = atan(x);
	d = (I*log(2)/2-I*log(1+I*x)/2) + (x-I)/4 + I*pow(x-I,2)/16 + Order(pow(x-I,3));
	result += check_series(e,I,d,3);
	
	// NB: here, at -I, Mathematica disagrees, but it is wrong -- they
	// pick up a complex phase by incorrectly expanding logarithms.
	e = atan(x);
	d = (-I*log(2)/2+I*log(1-I*x)/2) + (x+I)/4 - I*pow(x+I,2)/16 + Order(pow(x+I,3));
	result += check_series(e,-I,d,3);
	
	// This is basically the same as above, the branch point is at +/-1:
	e = atanh(x);
	d = (-log(2)/2+log(x+1)/2) + (x+1)/4 + pow(x+1,2)/16 + Order(pow(x+1,3));
	result += check_series(e,-1,d,3);
	
	return result;
}

// Test of the patch of Stefan Weinzierl that prevents an infinite loop if
// a factor in a product is a complicated way of writing zero.
static unsigned exam_series13()
{
	unsigned result = 0;

	ex e = (new mul(pow(2,x), (1/x*(-(1+x)/(1-x)) + (1+x)/x/(1-x)))
	       )->setflag(status_flags::evaluated);
	ex d = Order(x);
	result += check_series(e,0,d,1);

	return result;
}

// Test if (1+x)^(1/x) can be expanded.
static unsigned exam_series14()
{
	unsigned result = 0;

	ex e = pow(1+x, sin(x)/x);
	ex d = 1 + x - pow(x,3)/6 + Order(pow(x,4));
	try {
		result += check_series(e,0,d,4);
	} catch (const pole_error& err) {
		clog << "series expansion of " << e << " at 0 raised an exception." << endl;
		++result;
	}

	return result;
}

// Test expansion of powers of polynomials.
static unsigned exam_series15()
{
	unsigned result = 0;

	ex e = pow(x + pow(x,2), 2);

	result += check_series(e, 0, Order(1), 0);
	result += check_series(e, 0, Order(x), 1);
	result += check_series(e, 0, Order(pow(x,2)), 2);
	result += check_series(e, 0, pow(x,2) + Order(pow(x,3)), 3);
	result += check_series(e, 0, pow(x,2) + 2*pow(x,3) + Order(pow(x,4)), 4);
	result += check_series(e, 0, pow(x,2) + 2*pow(x,3) + pow(x,4), 5);
	result += check_series(e, 0, pow(x,2) + 2*pow(x,3) + pow(x,4), 6);

	return result;
}

unsigned exam_pseries()
{
	unsigned result = 0;
	
	cout << "examining series expansion" << flush;
	
	result += exam_series1();  cout << '.' << flush;
	result += exam_series2();  cout << '.' << flush;
	result += exam_series3();  cout << '.' << flush;
	result += exam_series4();  cout << '.' << flush;
	result += exam_series5();  cout << '.' << flush;
	result += exam_series6();  cout << '.' << flush;
	result += exam_series7();  cout << '.' << flush;
	result += exam_series8();  cout << '.' << flush;
	result += exam_series9();  cout << '.' << flush;
	result += exam_series10();  cout << '.' << flush;
	result += exam_series11();  cout << '.' << flush;
	result += exam_series12();  cout << '.' << flush;
	result += exam_series13();  cout << '.' << flush;
	result += exam_series14();  cout << '.' << flush;
	result += exam_series15();  cout << '.' << flush;
	
	return result;
}

int main(int argc, char** argv)
{
	return exam_pseries();
}