5 * Copyright (C) 2003 - 2007 Michael C. Ring
7 * Permission to use, copy, and distribute this software and its
8 * documentation for any purpose with or without fee is hereby granted,
9 * provided that the above copyright notice appear in all copies and
10 * that both that copyright notice and this permission notice appear
11 * in supporting documentation.
13 * Permission to modify the software is granted. Permission to distribute
14 * the modified code is granted. Modifications are to be distributed by
15 * using the file 'license.txt' as a template to modify the file header.
16 * 'license.txt' is available in the official MAPM distribution.
18 * This software is provided "as is" without express or implied warranty.
22 * $Id: mapm_lg3.c,v 1.7 2007/12/03 01:42:59 mike Exp $
24 * This file contains the function to compute log(2), log(10),
25 * and 1/log(10) to the desired precision using an AGM algorithm.
27 * $Log: mapm_lg3.c,v $
28 * Revision 1.7 2007/12/03 01:42:59 mike
31 * Revision 1.6 2003/12/09 01:25:06 mike
32 * actually compute the first term of the AGM iteration instead
33 * of assuming the inputs a=1 and b=10^-N.
35 * Revision 1.5 2003/12/04 03:19:16 mike
36 * rearrange logic in AGM to be more straight-forward
38 * Revision 1.4 2003/05/01 22:04:37 mike
41 * Revision 1.3 2003/05/01 21:58:31 mike
44 * Revision 1.2 2003/03/30 22:14:58 mike
47 * Revision 1.1 2003/03/30 21:18:04 mike
54 * using the 'R' function (defined below) for 'N' decimal places :
58 * log(2) = R(1, 0.5 * 10 ) - R(1, 10 )
62 * log(10) = R(1, 0.1 * 10 ) - R(1, 10 )
68 * log(x) = R(1, 10 / x) - R(1, 10 )
71 * I found this on a web site which went into considerable detail
72 * on the history of log(2). This formula is algebraically identical
73 * to the formula specified in J. Borwein and P. Borwein's book
74 * "PI and the AGM". (reference algorithm 7.2)
77 /****************************************************************************/
79 * check if our local copy of log(2) & log(10) is precise
80 * enough for our purpose. if not, calculate them so it's
81 * as precise as desired, accurate to at least 'places'.
83 void M_check_log_places(int places
)
85 M_APM tmp6
, tmp7
, tmp8
, tmp9
;
90 if (dplaces
> MM_lc_log_digits
)
92 MM_lc_log_digits
= dplaces
+ 4;
94 tmp6
= M_get_stack_var();
95 tmp7
= M_get_stack_var();
96 tmp8
= M_get_stack_var();
97 tmp9
= M_get_stack_var();
99 dplaces
+= 6 + (int)log10((double)places
);
101 m_apm_copy(tmp7
, MM_One
);
102 tmp7
->m_apm_exponent
= -places
;
104 M_log_AGM_R_func(tmp8
, dplaces
, MM_One
, tmp7
);
106 m_apm_multiply(tmp6
, tmp7
, MM_0_5
);
108 M_log_AGM_R_func(tmp9
, dplaces
, MM_One
, tmp6
);
110 m_apm_subtract(MM_lc_log2
, tmp9
, tmp8
); /* log(2) */
112 tmp7
->m_apm_exponent
-= 1; /* divide by 10 */
114 M_log_AGM_R_func(tmp9
, dplaces
, MM_One
, tmp7
);
116 m_apm_subtract(MM_lc_log10
, tmp9
, tmp8
); /* log(10) */
117 m_apm_reciprocal(MM_lc_log10R
, dplaces
, MM_lc_log10
); /* 1 / log(10) */
122 /****************************************************************************/
125 * define a notation for a function 'R' :
130 * R (a0, b0) = ------------------------------
142 * where a, b are the classic AGM iteration :
154 * define a variable 'c' for more efficient computation :
157 * c = 0.5 * (a - b ) , c = a - b
162 /****************************************************************************/
163 void M_log_AGM_R_func(M_APM rr
, int places
, M_APM aa
, M_APM bb
)
165 M_APM tmp1
, tmp2
, tmp3
, tmp4
, tmpC2
, sum
, pow_2
, tmpA0
, tmpB0
;
166 int tolerance
, dplaces
;
168 tmpA0
= M_get_stack_var();
169 tmpB0
= M_get_stack_var();
170 tmpC2
= M_get_stack_var();
171 tmp1
= M_get_stack_var();
172 tmp2
= M_get_stack_var();
173 tmp3
= M_get_stack_var();
174 tmp4
= M_get_stack_var();
175 sum
= M_get_stack_var();
176 pow_2
= M_get_stack_var();
178 tolerance
= places
+ 8;
179 dplaces
= places
+ 16;
181 m_apm_copy(tmpA0
, aa
);
182 m_apm_copy(tmpB0
, bb
);
183 m_apm_copy(pow_2
, MM_0_5
);
185 m_apm_multiply(tmp1
, aa
, aa
); /* 0.5 * [ a ^ 2 - b ^ 2 ] */
186 m_apm_multiply(tmp2
, bb
, bb
);
187 m_apm_subtract(tmp3
, tmp1
, tmp2
);
188 m_apm_multiply(sum
, MM_0_5
, tmp3
);
192 m_apm_subtract(tmp1
, tmpA0
, tmpB0
); /* C n+1 = 0.5 * [ An - Bn ] */
193 m_apm_multiply(tmp4
, MM_0_5
, tmp1
); /* C n+1 */
194 m_apm_multiply(tmpC2
, tmp4
, tmp4
); /* C n+1 ^ 2 */
198 m_apm_add(tmp1
, tmpA0
, tmpB0
);
199 m_apm_multiply(tmp3
, MM_0_5
, tmp1
);
201 m_apm_multiply(tmp2
, tmpA0
, tmpB0
);
202 m_apm_sqrt(tmpB0
, dplaces
, tmp2
);
204 m_apm_round(tmpA0
, dplaces
, tmp3
);
208 m_apm_multiply(tmp2
, MM_Two
, pow_2
);
209 m_apm_copy(pow_2
, tmp2
);
211 m_apm_multiply(tmp1
, tmpC2
, pow_2
);
212 m_apm_add(tmp3
, sum
, tmp1
);
214 if ((tmp1
->m_apm_sign
== 0) ||
215 ((-2 * tmp1
->m_apm_exponent
) > tolerance
))
218 m_apm_round(sum
, dplaces
, tmp3
);
221 m_apm_subtract(tmp4
, MM_One
, tmp3
);
222 m_apm_reciprocal(rr
, places
, tmp4
);
226 /****************************************************************************/