File: euler_number_vec.c

package info (click to toggle)
flint 2.4.4-2
  • links: PTS, VCS
  • area: main
  • in suites: jessie, jessie-kfreebsd
  • size: 21,904 kB
  • ctags: 13,326
  • sloc: ansic: 208,848; cpp: 11,358; sh: 564; makefile: 250
file content (147 lines) | stat: -rw-r--r-- 4,417 bytes parent folder | download | duplicates (4)
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
/*=============================================================================

    This file is part of FLINT.

    FLINT 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.

    FLINT 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 FLINT; if not, write to the Free Software
    Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301 USA

=============================================================================*/
/******************************************************************************

    Copyright (C) 2011 Fredrik Johansson

******************************************************************************/

#include <math.h>
#include "arith.h"

/* Computes length-m vector containing |E_{2k}| */
static void
__euler_number_vec_mod_p(mp_ptr res, mp_ptr tmp, slong m, nmod_t mod)
{
    mp_limb_t fac, c;
    slong k;

    /* Divide by factorials */
    fac = n_factorial_mod2_preinv(2*(m-1), mod.n, mod.ninv);
    c = n_invmod(fac, mod.n);
    for (k = m - 1; k >= 0; k--)
    {
        tmp[k] = c;
        c = n_mulmod2_preinv(c, (2*k)*(2*k-1), mod.n, mod.ninv);
    }

    _nmod_poly_inv_series(res, tmp, m, mod);

    /* Multiply by factorials */
    c = UWORD(1);
    for (k = 0; k < m; k++)
    {
        res[k] = n_mulmod2_preinv(res[k], c, mod.n, mod.ninv);
        c = n_mulmod2_preinv(c, (2*k+1)*(2*k+2), mod.n, mod.ninv);
        c = n_negmod(c, mod.n);
    }
}

#define CRT_MAX_RESOLUTION 16

void __euler_number_vec_multi_mod(fmpz * res, slong n)
{
    fmpz_comb_t comb[CRT_MAX_RESOLUTION];
    fmpz_comb_temp_t temp[CRT_MAX_RESOLUTION];
    mp_limb_t * primes;
    mp_limb_t * residues;
    mp_ptr * polys;
    mp_ptr temppoly;
    nmod_t mod;
    slong i, j, k, m, num_primes, num_primes_k, resolution;
    mp_bitcnt_t size, prime_bits;

    if (n < 1)
        return;

    /* Number of nonzero entries */
    m = (n + 1) / 2;

    resolution = FLINT_MAX(1, FLINT_MIN(CRT_MAX_RESOLUTION, m / 16));

    size = arith_euler_number_size(n);
    prime_bits = FLINT_BITS - 1;
    num_primes = (size + prime_bits - 1) / prime_bits;

    primes = flint_malloc(num_primes * sizeof(mp_limb_t));
    residues = flint_malloc(num_primes * sizeof(mp_limb_t));
    polys = flint_malloc(num_primes * sizeof(mp_ptr));

    /* Compute Euler numbers mod p */
    primes[0] = n_nextprime(UWORD(1)<<prime_bits, 0);
    for (k = 1; k < num_primes; k++)
        primes[k] = n_nextprime(primes[k-1], 0);
    temppoly = _nmod_vec_init(m);
    for (k = 0; k < num_primes; k++)
    {
        polys[k] = _nmod_vec_init(m);
        nmod_init(&mod, primes[k]);
        __euler_number_vec_mod_p(polys[k], temppoly, m, mod);
    }

    /* Init CRT comb */
    for (i = 0; i < resolution; i++)
    {
        fmpz_comb_init(comb[i], primes, num_primes * (i + 1) / resolution);
        fmpz_comb_temp_init(temp[i], comb[i]);
    }

    /* Trivial entries */
    for (k = 1; k < n; k += 2)
        fmpz_zero(res + k);

    /* Reconstruction */
    for (k = 0; k < n; k += 2)
    {
        size = arith_euler_number_size(k);
        /* Use only as large a comb as needed */
        num_primes_k = (size + prime_bits - 1) / prime_bits;
        for (i = 0; i < resolution; i++)
        {
            if (comb[i]->num_primes >= num_primes_k)
                break;
        }
        num_primes_k = comb[i]->num_primes;
        for (j = 0; j < num_primes_k; j++)
            residues[j] = polys[j][k / 2];
        fmpz_multi_CRT_ui(res + k, residues, comb[i], temp[i], 0);
        if (k % 4)
            fmpz_neg(res + k, res + k);
    }

    /* Cleanup */
    for (k = 0; k < num_primes; k++)
        _nmod_vec_clear(polys[k]);
    _nmod_vec_clear(temppoly);
    for (i = 0; i < resolution; i++)
    {
        fmpz_comb_temp_clear(temp[i]);
        fmpz_comb_clear(comb[i]);
    }

    flint_free(primes);
    flint_free(residues);
    flint_free(polys);
}

void arith_euler_number_vec(fmpz * res, slong n)
{
    __euler_number_vec_multi_mod(res, n);
}