1042 lines
30 KiB
C
1042 lines
30 KiB
C
#include <math.h>
|
|
#include <stdlib.h>
|
|
#include <string.h>
|
|
#include <stdio.h>
|
|
#include <complex.h>
|
|
#ifdef complex
|
|
#undef complex
|
|
#endif
|
|
#ifdef I
|
|
#undef I
|
|
#endif
|
|
|
|
#if defined(_WIN64)
|
|
typedef long long BLASLONG;
|
|
typedef unsigned long long BLASULONG;
|
|
#else
|
|
typedef long BLASLONG;
|
|
typedef unsigned long BLASULONG;
|
|
#endif
|
|
|
|
#ifdef LAPACK_ILP64
|
|
typedef BLASLONG blasint;
|
|
#if defined(_WIN64)
|
|
#define blasabs(x) llabs(x)
|
|
#else
|
|
#define blasabs(x) labs(x)
|
|
#endif
|
|
#else
|
|
typedef int blasint;
|
|
#define blasabs(x) abs(x)
|
|
#endif
|
|
|
|
typedef blasint integer;
|
|
|
|
typedef unsigned int uinteger;
|
|
typedef char *address;
|
|
typedef short int shortint;
|
|
typedef float real;
|
|
typedef double doublereal;
|
|
typedef struct { real r, i; } complex;
|
|
typedef struct { doublereal r, i; } doublecomplex;
|
|
#ifdef _MSC_VER
|
|
static inline _Fcomplex Cf(complex *z) {_Fcomplex zz={z->r , z->i}; return zz;}
|
|
static inline _Dcomplex Cd(doublecomplex *z) {_Dcomplex zz={z->r , z->i};return zz;}
|
|
static inline _Fcomplex * _pCf(complex *z) {return (_Fcomplex*)z;}
|
|
static inline _Dcomplex * _pCd(doublecomplex *z) {return (_Dcomplex*)z;}
|
|
#else
|
|
static inline _Complex float Cf(complex *z) {return z->r + z->i*_Complex_I;}
|
|
static inline _Complex double Cd(doublecomplex *z) {return z->r + z->i*_Complex_I;}
|
|
static inline _Complex float * _pCf(complex *z) {return (_Complex float*)z;}
|
|
static inline _Complex double * _pCd(doublecomplex *z) {return (_Complex double*)z;}
|
|
#endif
|
|
#define pCf(z) (*_pCf(z))
|
|
#define pCd(z) (*_pCd(z))
|
|
typedef blasint logical;
|
|
|
|
typedef char logical1;
|
|
typedef char integer1;
|
|
|
|
#define TRUE_ (1)
|
|
#define FALSE_ (0)
|
|
|
|
/* Extern is for use with -E */
|
|
#ifndef Extern
|
|
#define Extern extern
|
|
#endif
|
|
|
|
/* I/O stuff */
|
|
|
|
typedef int flag;
|
|
typedef int ftnlen;
|
|
typedef int ftnint;
|
|
|
|
/*external read, write*/
|
|
typedef struct
|
|
{ flag cierr;
|
|
ftnint ciunit;
|
|
flag ciend;
|
|
char *cifmt;
|
|
ftnint cirec;
|
|
} cilist;
|
|
|
|
/*internal read, write*/
|
|
typedef struct
|
|
{ flag icierr;
|
|
char *iciunit;
|
|
flag iciend;
|
|
char *icifmt;
|
|
ftnint icirlen;
|
|
ftnint icirnum;
|
|
} icilist;
|
|
|
|
/*open*/
|
|
typedef struct
|
|
{ flag oerr;
|
|
ftnint ounit;
|
|
char *ofnm;
|
|
ftnlen ofnmlen;
|
|
char *osta;
|
|
char *oacc;
|
|
char *ofm;
|
|
ftnint orl;
|
|
char *oblnk;
|
|
} olist;
|
|
|
|
/*close*/
|
|
typedef struct
|
|
{ flag cerr;
|
|
ftnint cunit;
|
|
char *csta;
|
|
} cllist;
|
|
|
|
/*rewind, backspace, endfile*/
|
|
typedef struct
|
|
{ flag aerr;
|
|
ftnint aunit;
|
|
} alist;
|
|
|
|
/* inquire */
|
|
typedef struct
|
|
{ flag inerr;
|
|
ftnint inunit;
|
|
char *infile;
|
|
ftnlen infilen;
|
|
ftnint *inex; /*parameters in standard's order*/
|
|
ftnint *inopen;
|
|
ftnint *innum;
|
|
ftnint *innamed;
|
|
char *inname;
|
|
ftnlen innamlen;
|
|
char *inacc;
|
|
ftnlen inacclen;
|
|
char *inseq;
|
|
ftnlen inseqlen;
|
|
char *indir;
|
|
ftnlen indirlen;
|
|
char *infmt;
|
|
ftnlen infmtlen;
|
|
char *inform;
|
|
ftnint informlen;
|
|
char *inunf;
|
|
ftnlen inunflen;
|
|
ftnint *inrecl;
|
|
ftnint *innrec;
|
|
char *inblank;
|
|
ftnlen inblanklen;
|
|
} inlist;
|
|
|
|
#define VOID void
|
|
|
|
union Multitype { /* for multiple entry points */
|
|
integer1 g;
|
|
shortint h;
|
|
integer i;
|
|
/* longint j; */
|
|
real r;
|
|
doublereal d;
|
|
complex c;
|
|
doublecomplex z;
|
|
};
|
|
|
|
typedef union Multitype Multitype;
|
|
|
|
struct Vardesc { /* for Namelist */
|
|
char *name;
|
|
char *addr;
|
|
ftnlen *dims;
|
|
int type;
|
|
};
|
|
typedef struct Vardesc Vardesc;
|
|
|
|
struct Namelist {
|
|
char *name;
|
|
Vardesc **vars;
|
|
int nvars;
|
|
};
|
|
typedef struct Namelist Namelist;
|
|
|
|
#define abs(x) ((x) >= 0 ? (x) : -(x))
|
|
#define dabs(x) (fabs(x))
|
|
#define f2cmin(a,b) ((a) <= (b) ? (a) : (b))
|
|
#define f2cmax(a,b) ((a) >= (b) ? (a) : (b))
|
|
#define dmin(a,b) (f2cmin(a,b))
|
|
#define dmax(a,b) (f2cmax(a,b))
|
|
#define bit_test(a,b) ((a) >> (b) & 1)
|
|
#define bit_clear(a,b) ((a) & ~((uinteger)1 << (b)))
|
|
#define bit_set(a,b) ((a) | ((uinteger)1 << (b)))
|
|
|
|
#define abort_() { sig_die("Fortran abort routine called", 1); }
|
|
#define c_abs(z) (cabsf(Cf(z)))
|
|
#define c_cos(R,Z) { pCf(R)=ccos(Cf(Z)); }
|
|
#ifdef _MSC_VER
|
|
#define c_div(c, a, b) {Cf(c)._Val[0] = (Cf(a)._Val[0]/Cf(b)._Val[0]); Cf(c)._Val[1]=(Cf(a)._Val[1]/Cf(b)._Val[1]);}
|
|
#define z_div(c, a, b) {Cd(c)._Val[0] = (Cd(a)._Val[0]/Cd(b)._Val[0]); Cd(c)._Val[1]=(Cd(a)._Val[1]/df(b)._Val[1]);}
|
|
#else
|
|
#define c_div(c, a, b) {pCf(c) = Cf(a)/Cf(b);}
|
|
#define z_div(c, a, b) {pCd(c) = Cd(a)/Cd(b);}
|
|
#endif
|
|
#define c_exp(R, Z) {pCf(R) = cexpf(Cf(Z));}
|
|
#define c_log(R, Z) {pCf(R) = clogf(Cf(Z));}
|
|
#define c_sin(R, Z) {pCf(R) = csinf(Cf(Z));}
|
|
//#define c_sqrt(R, Z) {*(R) = csqrtf(Cf(Z));}
|
|
#define c_sqrt(R, Z) {pCf(R) = csqrtf(Cf(Z));}
|
|
#define d_abs(x) (fabs(*(x)))
|
|
#define d_acos(x) (acos(*(x)))
|
|
#define d_asin(x) (asin(*(x)))
|
|
#define d_atan(x) (atan(*(x)))
|
|
#define d_atn2(x, y) (atan2(*(x),*(y)))
|
|
#define d_cnjg(R, Z) { pCd(R) = conj(Cd(Z)); }
|
|
#define r_cnjg(R, Z) { pCf(R) = conjf(Cf(Z)); }
|
|
#define d_cos(x) (cos(*(x)))
|
|
#define d_cosh(x) (cosh(*(x)))
|
|
#define d_dim(__a, __b) ( *(__a) > *(__b) ? *(__a) - *(__b) : 0.0 )
|
|
#define d_exp(x) (exp(*(x)))
|
|
#define d_imag(z) (cimag(Cd(z)))
|
|
#define r_imag(z) (cimagf(Cf(z)))
|
|
#define d_int(__x) (*(__x)>0 ? floor(*(__x)) : -floor(- *(__x)))
|
|
#define r_int(__x) (*(__x)>0 ? floor(*(__x)) : -floor(- *(__x)))
|
|
#define d_lg10(x) ( 0.43429448190325182765 * log(*(x)) )
|
|
#define r_lg10(x) ( 0.43429448190325182765 * log(*(x)) )
|
|
#define d_log(x) (log(*(x)))
|
|
#define d_mod(x, y) (fmod(*(x), *(y)))
|
|
#define u_nint(__x) ((__x)>=0 ? floor((__x) + .5) : -floor(.5 - (__x)))
|
|
#define d_nint(x) u_nint(*(x))
|
|
#define u_sign(__a,__b) ((__b) >= 0 ? ((__a) >= 0 ? (__a) : -(__a)) : -((__a) >= 0 ? (__a) : -(__a)))
|
|
#define d_sign(a,b) u_sign(*(a),*(b))
|
|
#define r_sign(a,b) u_sign(*(a),*(b))
|
|
#define d_sin(x) (sin(*(x)))
|
|
#define d_sinh(x) (sinh(*(x)))
|
|
#define d_sqrt(x) (sqrt(*(x)))
|
|
#define d_tan(x) (tan(*(x)))
|
|
#define d_tanh(x) (tanh(*(x)))
|
|
#define i_abs(x) abs(*(x))
|
|
#define i_dnnt(x) ((integer)u_nint(*(x)))
|
|
#define i_len(s, n) (n)
|
|
#define i_nint(x) ((integer)u_nint(*(x)))
|
|
#define i_sign(a,b) ((integer)u_sign((integer)*(a),(integer)*(b)))
|
|
#define pow_dd(ap, bp) ( pow(*(ap), *(bp)))
|
|
#define pow_si(B,E) spow_ui(*(B),*(E))
|
|
#define pow_ri(B,E) spow_ui(*(B),*(E))
|
|
#define pow_di(B,E) dpow_ui(*(B),*(E))
|
|
#define pow_zi(p, a, b) {pCd(p) = zpow_ui(Cd(a), *(b));}
|
|
#define pow_ci(p, a, b) {pCf(p) = cpow_ui(Cf(a), *(b));}
|
|
#define pow_zz(R,A,B) {pCd(R) = cpow(Cd(A),*(B));}
|
|
#define s_cat(lpp, rpp, rnp, np, llp) { ftnlen i, nc, ll; char *f__rp, *lp; ll = (llp); lp = (lpp); for(i=0; i < (int)*(np); ++i) { nc = ll; if((rnp)[i] < nc) nc = (rnp)[i]; ll -= nc; f__rp = (rpp)[i]; while(--nc >= 0) *lp++ = *(f__rp)++; } while(--ll >= 0) *lp++ = ' '; }
|
|
#define s_cmp(a,b,c,d) ((integer)strncmp((a),(b),f2cmin((c),(d))))
|
|
#define s_copy(A,B,C,D) { int __i,__m; for (__i=0, __m=f2cmin((C),(D)); __i<__m && (B)[__i] != 0; ++__i) (A)[__i] = (B)[__i]; }
|
|
#define sig_die(s, kill) { exit(1); }
|
|
#define s_stop(s, n) {exit(0);}
|
|
static char junk[] = "\n@(#)LIBF77 VERSION 19990503\n";
|
|
#define z_abs(z) (cabs(Cd(z)))
|
|
#define z_exp(R, Z) {pCd(R) = cexp(Cd(Z));}
|
|
#define z_sqrt(R, Z) {pCd(R) = csqrt(Cd(Z));}
|
|
#define myexit_() break;
|
|
#define mycycle() continue;
|
|
#define myceiling(w) {ceil(w)}
|
|
#define myhuge(w) {HUGE_VAL}
|
|
//#define mymaxloc_(w,s,e,n) {if (sizeof(*(w)) == sizeof(double)) dmaxloc_((w),*(s),*(e),n); else dmaxloc_((w),*(s),*(e),n);}
|
|
#define mymaxloc(w,s,e,n) {dmaxloc_(w,*(s),*(e),n)}
|
|
|
|
/* procedure parameter types for -A and -C++ */
|
|
|
|
|
|
#ifdef __cplusplus
|
|
typedef logical (*L_fp)(...);
|
|
#else
|
|
typedef logical (*L_fp)();
|
|
#endif
|
|
|
|
static float spow_ui(float x, integer n) {
|
|
float pow=1.0; unsigned long int u;
|
|
if(n != 0) {
|
|
if(n < 0) n = -n, x = 1/x;
|
|
for(u = n; ; ) {
|
|
if(u & 01) pow *= x;
|
|
if(u >>= 1) x *= x;
|
|
else break;
|
|
}
|
|
}
|
|
return pow;
|
|
}
|
|
static double dpow_ui(double x, integer n) {
|
|
double pow=1.0; unsigned long int u;
|
|
if(n != 0) {
|
|
if(n < 0) n = -n, x = 1/x;
|
|
for(u = n; ; ) {
|
|
if(u & 01) pow *= x;
|
|
if(u >>= 1) x *= x;
|
|
else break;
|
|
}
|
|
}
|
|
return pow;
|
|
}
|
|
#ifdef _MSC_VER
|
|
static _Fcomplex cpow_ui(complex x, integer n) {
|
|
complex pow={1.0,0.0}; unsigned long int u;
|
|
if(n != 0) {
|
|
if(n < 0) n = -n, x.r = 1/x.r, x.i=1/x.i;
|
|
for(u = n; ; ) {
|
|
if(u & 01) pow.r *= x.r, pow.i *= x.i;
|
|
if(u >>= 1) x.r *= x.r, x.i *= x.i;
|
|
else break;
|
|
}
|
|
}
|
|
_Fcomplex p={pow.r, pow.i};
|
|
return p;
|
|
}
|
|
#else
|
|
static _Complex float cpow_ui(_Complex float x, integer n) {
|
|
_Complex float pow=1.0; unsigned long int u;
|
|
if(n != 0) {
|
|
if(n < 0) n = -n, x = 1/x;
|
|
for(u = n; ; ) {
|
|
if(u & 01) pow *= x;
|
|
if(u >>= 1) x *= x;
|
|
else break;
|
|
}
|
|
}
|
|
return pow;
|
|
}
|
|
#endif
|
|
#ifdef _MSC_VER
|
|
static _Dcomplex zpow_ui(_Dcomplex x, integer n) {
|
|
_Dcomplex pow={1.0,0.0}; unsigned long int u;
|
|
if(n != 0) {
|
|
if(n < 0) n = -n, x._Val[0] = 1/x._Val[0], x._Val[1] =1/x._Val[1];
|
|
for(u = n; ; ) {
|
|
if(u & 01) pow._Val[0] *= x._Val[0], pow._Val[1] *= x._Val[1];
|
|
if(u >>= 1) x._Val[0] *= x._Val[0], x._Val[1] *= x._Val[1];
|
|
else break;
|
|
}
|
|
}
|
|
_Dcomplex p = {pow._Val[0], pow._Val[1]};
|
|
return p;
|
|
}
|
|
#else
|
|
static _Complex double zpow_ui(_Complex double x, integer n) {
|
|
_Complex double pow=1.0; unsigned long int u;
|
|
if(n != 0) {
|
|
if(n < 0) n = -n, x = 1/x;
|
|
for(u = n; ; ) {
|
|
if(u & 01) pow *= x;
|
|
if(u >>= 1) x *= x;
|
|
else break;
|
|
}
|
|
}
|
|
return pow;
|
|
}
|
|
#endif
|
|
static integer pow_ii(integer x, integer n) {
|
|
integer pow; unsigned long int u;
|
|
if (n <= 0) {
|
|
if (n == 0 || x == 1) pow = 1;
|
|
else if (x != -1) pow = x == 0 ? 1/x : 0;
|
|
else n = -n;
|
|
}
|
|
if ((n > 0) || !(n == 0 || x == 1 || x != -1)) {
|
|
u = n;
|
|
for(pow = 1; ; ) {
|
|
if(u & 01) pow *= x;
|
|
if(u >>= 1) x *= x;
|
|
else break;
|
|
}
|
|
}
|
|
return pow;
|
|
}
|
|
static integer dmaxloc_(double *w, integer s, integer e, integer *n)
|
|
{
|
|
double m; integer i, mi;
|
|
for(m=w[s-1], mi=s, i=s+1; i<=e; i++)
|
|
if (w[i-1]>m) mi=i ,m=w[i-1];
|
|
return mi-s+1;
|
|
}
|
|
static integer smaxloc_(float *w, integer s, integer e, integer *n)
|
|
{
|
|
float m; integer i, mi;
|
|
for(m=w[s-1], mi=s, i=s+1; i<=e; i++)
|
|
if (w[i-1]>m) mi=i ,m=w[i-1];
|
|
return mi-s+1;
|
|
}
|
|
static inline void cdotc_(complex *z, integer *n_, complex *x, integer *incx_, complex *y, integer *incy_) {
|
|
integer n = *n_, incx = *incx_, incy = *incy_, i;
|
|
#ifdef _MSC_VER
|
|
_Fcomplex zdotc = {0.0, 0.0};
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += conjf(Cf(&x[i]))._Val[0] * Cf(&y[i])._Val[0];
|
|
zdotc._Val[1] += conjf(Cf(&x[i]))._Val[1] * Cf(&y[i])._Val[1];
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += conjf(Cf(&x[i*incx]))._Val[0] * Cf(&y[i*incy])._Val[0];
|
|
zdotc._Val[1] += conjf(Cf(&x[i*incx]))._Val[1] * Cf(&y[i*incy])._Val[1];
|
|
}
|
|
}
|
|
pCf(z) = zdotc;
|
|
}
|
|
#else
|
|
_Complex float zdotc = 0.0;
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += conjf(Cf(&x[i])) * Cf(&y[i]);
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += conjf(Cf(&x[i*incx])) * Cf(&y[i*incy]);
|
|
}
|
|
}
|
|
pCf(z) = zdotc;
|
|
}
|
|
#endif
|
|
static inline void zdotc_(doublecomplex *z, integer *n_, doublecomplex *x, integer *incx_, doublecomplex *y, integer *incy_) {
|
|
integer n = *n_, incx = *incx_, incy = *incy_, i;
|
|
#ifdef _MSC_VER
|
|
_Dcomplex zdotc = {0.0, 0.0};
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += conj(Cd(&x[i]))._Val[0] * Cd(&y[i])._Val[0];
|
|
zdotc._Val[1] += conj(Cd(&x[i]))._Val[1] * Cd(&y[i])._Val[1];
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += conj(Cd(&x[i*incx]))._Val[0] * Cd(&y[i*incy])._Val[0];
|
|
zdotc._Val[1] += conj(Cd(&x[i*incx]))._Val[1] * Cd(&y[i*incy])._Val[1];
|
|
}
|
|
}
|
|
pCd(z) = zdotc;
|
|
}
|
|
#else
|
|
_Complex double zdotc = 0.0;
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += conj(Cd(&x[i])) * Cd(&y[i]);
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += conj(Cd(&x[i*incx])) * Cd(&y[i*incy]);
|
|
}
|
|
}
|
|
pCd(z) = zdotc;
|
|
}
|
|
#endif
|
|
static inline void cdotu_(complex *z, integer *n_, complex *x, integer *incx_, complex *y, integer *incy_) {
|
|
integer n = *n_, incx = *incx_, incy = *incy_, i;
|
|
#ifdef _MSC_VER
|
|
_Fcomplex zdotc = {0.0, 0.0};
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += Cf(&x[i])._Val[0] * Cf(&y[i])._Val[0];
|
|
zdotc._Val[1] += Cf(&x[i])._Val[1] * Cf(&y[i])._Val[1];
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += Cf(&x[i*incx])._Val[0] * Cf(&y[i*incy])._Val[0];
|
|
zdotc._Val[1] += Cf(&x[i*incx])._Val[1] * Cf(&y[i*incy])._Val[1];
|
|
}
|
|
}
|
|
pCf(z) = zdotc;
|
|
}
|
|
#else
|
|
_Complex float zdotc = 0.0;
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += Cf(&x[i]) * Cf(&y[i]);
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += Cf(&x[i*incx]) * Cf(&y[i*incy]);
|
|
}
|
|
}
|
|
pCf(z) = zdotc;
|
|
}
|
|
#endif
|
|
static inline void zdotu_(doublecomplex *z, integer *n_, doublecomplex *x, integer *incx_, doublecomplex *y, integer *incy_) {
|
|
integer n = *n_, incx = *incx_, incy = *incy_, i;
|
|
#ifdef _MSC_VER
|
|
_Dcomplex zdotc = {0.0, 0.0};
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += Cd(&x[i])._Val[0] * Cd(&y[i])._Val[0];
|
|
zdotc._Val[1] += Cd(&x[i])._Val[1] * Cd(&y[i])._Val[1];
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc._Val[0] += Cd(&x[i*incx])._Val[0] * Cd(&y[i*incy])._Val[0];
|
|
zdotc._Val[1] += Cd(&x[i*incx])._Val[1] * Cd(&y[i*incy])._Val[1];
|
|
}
|
|
}
|
|
pCd(z) = zdotc;
|
|
}
|
|
#else
|
|
_Complex double zdotc = 0.0;
|
|
if (incx == 1 && incy == 1) {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += Cd(&x[i]) * Cd(&y[i]);
|
|
}
|
|
} else {
|
|
for (i=0;i<n;i++) { /* zdotc = zdotc + dconjg(x(i))* y(i) */
|
|
zdotc += Cd(&x[i*incx]) * Cd(&y[i*incy]);
|
|
}
|
|
}
|
|
pCd(z) = zdotc;
|
|
}
|
|
#endif
|
|
/* -- translated by f2c (version 20000121).
|
|
You must link the resulting object file with the libraries:
|
|
-lf2c -lm (in that order)
|
|
*/
|
|
|
|
|
|
|
|
|
|
/* Table of constant values */
|
|
|
|
static integer c__9 = 9;
|
|
static integer c__0 = 0;
|
|
static integer c__2 = 2;
|
|
static doublereal c_b17 = 0.;
|
|
static doublereal c_b18 = 1.;
|
|
static integer c__1 = 1;
|
|
|
|
/* > \brief \b DSTEDC */
|
|
|
|
/* =========== DOCUMENTATION =========== */
|
|
|
|
/* Online html documentation available at */
|
|
/* http://www.netlib.org/lapack/explore-html/ */
|
|
|
|
/* > \htmlonly */
|
|
/* > Download DSTEDC + dependencies */
|
|
/* > <a href="http://www.netlib.org/cgi-bin/netlibfiles.tgz?format=tgz&filename=/lapack/lapack_routine/dstedc.
|
|
f"> */
|
|
/* > [TGZ]</a> */
|
|
/* > <a href="http://www.netlib.org/cgi-bin/netlibfiles.zip?format=zip&filename=/lapack/lapack_routine/dstedc.
|
|
f"> */
|
|
/* > [ZIP]</a> */
|
|
/* > <a href="http://www.netlib.org/cgi-bin/netlibfiles.txt?format=txt&filename=/lapack/lapack_routine/dstedc.
|
|
f"> */
|
|
/* > [TXT]</a> */
|
|
/* > \endhtmlonly */
|
|
|
|
/* Definition: */
|
|
/* =========== */
|
|
|
|
/* SUBROUTINE DSTEDC( COMPZ, N, D, E, Z, LDZ, WORK, LWORK, IWORK, */
|
|
/* LIWORK, INFO ) */
|
|
|
|
/* CHARACTER COMPZ */
|
|
/* INTEGER INFO, LDZ, LIWORK, LWORK, N */
|
|
/* INTEGER IWORK( * ) */
|
|
/* DOUBLE PRECISION D( * ), E( * ), WORK( * ), Z( LDZ, * ) */
|
|
|
|
|
|
/* > \par Purpose: */
|
|
/* ============= */
|
|
/* > */
|
|
/* > \verbatim */
|
|
/* > */
|
|
/* > DSTEDC computes all eigenvalues and, optionally, eigenvectors of a */
|
|
/* > symmetric tridiagonal matrix using the divide and conquer method. */
|
|
/* > The eigenvectors of a full or band real symmetric matrix can also be */
|
|
/* > found if DSYTRD or DSPTRD or DSBTRD has been used to reduce this */
|
|
/* > matrix to tridiagonal form. */
|
|
/* > */
|
|
/* > This code makes very mild assumptions about floating point */
|
|
/* > arithmetic. It will work on machines with a guard digit in */
|
|
/* > add/subtract, or on those binary machines without guard digits */
|
|
/* > which subtract like the Cray X-MP, Cray Y-MP, Cray C-90, or Cray-2. */
|
|
/* > It could conceivably fail on hexadecimal or decimal machines */
|
|
/* > without guard digits, but we know of none. See DLAED3 for details. */
|
|
/* > \endverbatim */
|
|
|
|
/* Arguments: */
|
|
/* ========== */
|
|
|
|
/* > \param[in] COMPZ */
|
|
/* > \verbatim */
|
|
/* > COMPZ is CHARACTER*1 */
|
|
/* > = 'N': Compute eigenvalues only. */
|
|
/* > = 'I': Compute eigenvectors of tridiagonal matrix also. */
|
|
/* > = 'V': Compute eigenvectors of original dense symmetric */
|
|
/* > matrix also. On entry, Z contains the orthogonal */
|
|
/* > matrix used to reduce the original matrix to */
|
|
/* > tridiagonal form. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[in] N */
|
|
/* > \verbatim */
|
|
/* > N is INTEGER */
|
|
/* > The dimension of the symmetric tridiagonal matrix. N >= 0. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[in,out] D */
|
|
/* > \verbatim */
|
|
/* > D is DOUBLE PRECISION array, dimension (N) */
|
|
/* > On entry, the diagonal elements of the tridiagonal matrix. */
|
|
/* > On exit, if INFO = 0, the eigenvalues in ascending order. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[in,out] E */
|
|
/* > \verbatim */
|
|
/* > E is DOUBLE PRECISION array, dimension (N-1) */
|
|
/* > On entry, the subdiagonal elements of the tridiagonal matrix. */
|
|
/* > On exit, E has been destroyed. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[in,out] Z */
|
|
/* > \verbatim */
|
|
/* > Z is DOUBLE PRECISION array, dimension (LDZ,N) */
|
|
/* > On entry, if COMPZ = 'V', then Z contains the orthogonal */
|
|
/* > matrix used in the reduction to tridiagonal form. */
|
|
/* > On exit, if INFO = 0, then if COMPZ = 'V', Z contains the */
|
|
/* > orthonormal eigenvectors of the original symmetric matrix, */
|
|
/* > and if COMPZ = 'I', Z contains the orthonormal eigenvectors */
|
|
/* > of the symmetric tridiagonal matrix. */
|
|
/* > If COMPZ = 'N', then Z is not referenced. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[in] LDZ */
|
|
/* > \verbatim */
|
|
/* > LDZ is INTEGER */
|
|
/* > The leading dimension of the array Z. LDZ >= 1. */
|
|
/* > If eigenvectors are desired, then LDZ >= f2cmax(1,N). */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[out] WORK */
|
|
/* > \verbatim */
|
|
/* > WORK is DOUBLE PRECISION array, dimension (MAX(1,LWORK)) */
|
|
/* > On exit, if INFO = 0, WORK(1) returns the optimal LWORK. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[in] LWORK */
|
|
/* > \verbatim */
|
|
/* > LWORK is INTEGER */
|
|
/* > The dimension of the array WORK. */
|
|
/* > If COMPZ = 'N' or N <= 1 then LWORK must be at least 1. */
|
|
/* > If COMPZ = 'V' and N > 1 then LWORK must be at least */
|
|
/* > ( 1 + 3*N + 2*N*lg N + 4*N**2 ), */
|
|
/* > where lg( N ) = smallest integer k such */
|
|
/* > that 2**k >= N. */
|
|
/* > If COMPZ = 'I' and N > 1 then LWORK must be at least */
|
|
/* > ( 1 + 4*N + N**2 ). */
|
|
/* > Note that for COMPZ = 'I' or 'V', then if N is less than or */
|
|
/* > equal to the minimum divide size, usually 25, then LWORK need */
|
|
/* > only be f2cmax(1,2*(N-1)). */
|
|
/* > */
|
|
/* > If LWORK = -1, then a workspace query is assumed; the routine */
|
|
/* > only calculates the optimal size of the WORK array, returns */
|
|
/* > this value as the first entry of the WORK array, and no error */
|
|
/* > message related to LWORK is issued by XERBLA. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[out] IWORK */
|
|
/* > \verbatim */
|
|
/* > IWORK is INTEGER array, dimension (MAX(1,LIWORK)) */
|
|
/* > On exit, if INFO = 0, IWORK(1) returns the optimal LIWORK. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[in] LIWORK */
|
|
/* > \verbatim */
|
|
/* > LIWORK is INTEGER */
|
|
/* > The dimension of the array IWORK. */
|
|
/* > If COMPZ = 'N' or N <= 1 then LIWORK must be at least 1. */
|
|
/* > If COMPZ = 'V' and N > 1 then LIWORK must be at least */
|
|
/* > ( 6 + 6*N + 5*N*lg N ). */
|
|
/* > If COMPZ = 'I' and N > 1 then LIWORK must be at least */
|
|
/* > ( 3 + 5*N ). */
|
|
/* > Note that for COMPZ = 'I' or 'V', then if N is less than or */
|
|
/* > equal to the minimum divide size, usually 25, then LIWORK */
|
|
/* > need only be 1. */
|
|
/* > */
|
|
/* > If LIWORK = -1, then a workspace query is assumed; the */
|
|
/* > routine only calculates the optimal size of the IWORK array, */
|
|
/* > returns this value as the first entry of the IWORK array, and */
|
|
/* > no error message related to LIWORK is issued by XERBLA. */
|
|
/* > \endverbatim */
|
|
/* > */
|
|
/* > \param[out] INFO */
|
|
/* > \verbatim */
|
|
/* > INFO is INTEGER */
|
|
/* > = 0: successful exit. */
|
|
/* > < 0: if INFO = -i, the i-th argument had an illegal value. */
|
|
/* > > 0: The algorithm failed to compute an eigenvalue while */
|
|
/* > working on the submatrix lying in rows and columns */
|
|
/* > INFO/(N+1) through mod(INFO,N+1). */
|
|
/* > \endverbatim */
|
|
|
|
/* Authors: */
|
|
/* ======== */
|
|
|
|
/* > \author Univ. of Tennessee */
|
|
/* > \author Univ. of California Berkeley */
|
|
/* > \author Univ. of Colorado Denver */
|
|
/* > \author NAG Ltd. */
|
|
|
|
/* > \date June 2017 */
|
|
|
|
/* > \ingroup auxOTHERcomputational */
|
|
|
|
/* > \par Contributors: */
|
|
/* ================== */
|
|
/* > */
|
|
/* > Jeff Rutter, Computer Science Division, University of California */
|
|
/* > at Berkeley, USA \n */
|
|
/* > Modified by Francoise Tisseur, University of Tennessee */
|
|
/* > */
|
|
/* ===================================================================== */
|
|
/* Subroutine */ void dstedc_(char *compz, integer *n, doublereal *d__,
|
|
doublereal *e, doublereal *z__, integer *ldz, doublereal *work,
|
|
integer *lwork, integer *iwork, integer *liwork, integer *info)
|
|
{
|
|
/* System generated locals */
|
|
integer z_dim1, z_offset, i__1, i__2;
|
|
doublereal d__1, d__2;
|
|
|
|
/* Local variables */
|
|
doublereal tiny;
|
|
integer i__, j, k, m;
|
|
doublereal p;
|
|
extern /* Subroutine */ void dgemm_(char *, char *, integer *, integer *,
|
|
integer *, doublereal *, doublereal *, integer *, doublereal *,
|
|
integer *, doublereal *, doublereal *, integer *);
|
|
extern logical lsame_(char *, char *);
|
|
extern /* Subroutine */ void dswap_(integer *, doublereal *, integer *,
|
|
doublereal *, integer *);
|
|
integer lwmin;
|
|
extern /* Subroutine */ void dlaed0_(integer *, integer *, integer *,
|
|
doublereal *, doublereal *, doublereal *, integer *, doublereal *,
|
|
integer *, doublereal *, integer *, integer *);
|
|
integer start, ii;
|
|
extern doublereal dlamch_(char *);
|
|
extern /* Subroutine */ void dlascl_(char *, integer *, integer *,
|
|
doublereal *, doublereal *, integer *, integer *, doublereal *,
|
|
integer *, integer *), dlacpy_(char *, integer *, integer
|
|
*, doublereal *, integer *, doublereal *, integer *),
|
|
dlaset_(char *, integer *, integer *, doublereal *, doublereal *,
|
|
doublereal *, integer *);
|
|
extern integer ilaenv_(integer *, char *, char *, integer *, integer *,
|
|
integer *, integer *, ftnlen, ftnlen);
|
|
extern /* Subroutine */ int xerbla_(char *, integer *, ftnlen);
|
|
integer finish;
|
|
extern doublereal dlanst_(char *, integer *, doublereal *, doublereal *);
|
|
extern /* Subroutine */ void dsterf_(integer *, doublereal *, doublereal *,
|
|
integer *), dlasrt_(char *, integer *, doublereal *, integer *);
|
|
integer liwmin, icompz;
|
|
extern /* Subroutine */ void dsteqr_(char *, integer *, doublereal *,
|
|
doublereal *, doublereal *, integer *, doublereal *, integer *);
|
|
doublereal orgnrm;
|
|
logical lquery;
|
|
integer smlsiz, storez, strtrw, lgn;
|
|
doublereal eps;
|
|
|
|
|
|
/* -- LAPACK computational routine (version 3.7.1) -- */
|
|
/* -- LAPACK is a software package provided by Univ. of Tennessee, -- */
|
|
/* -- Univ. of California Berkeley, Univ. of Colorado Denver and NAG Ltd..-- */
|
|
/* June 2017 */
|
|
|
|
|
|
/* ===================================================================== */
|
|
|
|
|
|
/* Test the input parameters. */
|
|
|
|
/* Parameter adjustments */
|
|
--d__;
|
|
--e;
|
|
z_dim1 = *ldz;
|
|
z_offset = 1 + z_dim1 * 1;
|
|
z__ -= z_offset;
|
|
--work;
|
|
--iwork;
|
|
|
|
/* Function Body */
|
|
*info = 0;
|
|
lquery = *lwork == -1 || *liwork == -1;
|
|
|
|
if (lsame_(compz, "N")) {
|
|
icompz = 0;
|
|
} else if (lsame_(compz, "V")) {
|
|
icompz = 1;
|
|
} else if (lsame_(compz, "I")) {
|
|
icompz = 2;
|
|
} else {
|
|
icompz = -1;
|
|
}
|
|
if (icompz < 0) {
|
|
*info = -1;
|
|
} else if (*n < 0) {
|
|
*info = -2;
|
|
} else if (*ldz < 1 || icompz > 0 && *ldz < f2cmax(1,*n)) {
|
|
*info = -6;
|
|
}
|
|
|
|
if (*info == 0) {
|
|
|
|
/* Compute the workspace requirements */
|
|
|
|
smlsiz = ilaenv_(&c__9, "DSTEDC", " ", &c__0, &c__0, &c__0, &c__0, (
|
|
ftnlen)6, (ftnlen)1);
|
|
if (*n <= 1 || icompz == 0) {
|
|
liwmin = 1;
|
|
lwmin = 1;
|
|
} else if (*n <= smlsiz) {
|
|
liwmin = 1;
|
|
lwmin = *n - 1 << 1;
|
|
} else {
|
|
lgn = (integer) (log((doublereal) (*n)) / log(2.));
|
|
if (pow_ii(c__2, lgn) < *n) {
|
|
++lgn;
|
|
}
|
|
if (pow_ii(c__2, lgn) < *n) {
|
|
++lgn;
|
|
}
|
|
if (icompz == 1) {
|
|
/* Computing 2nd power */
|
|
i__1 = *n;
|
|
lwmin = *n * 3 + 1 + (*n << 1) * lgn + (i__1 * i__1 << 2);
|
|
liwmin = *n * 6 + 6 + *n * 5 * lgn;
|
|
} else if (icompz == 2) {
|
|
/* Computing 2nd power */
|
|
i__1 = *n;
|
|
lwmin = (*n << 2) + 1 + i__1 * i__1;
|
|
liwmin = *n * 5 + 3;
|
|
}
|
|
}
|
|
work[1] = (doublereal) lwmin;
|
|
iwork[1] = liwmin;
|
|
|
|
if (*lwork < lwmin && ! lquery) {
|
|
*info = -8;
|
|
} else if (*liwork < liwmin && ! lquery) {
|
|
*info = -10;
|
|
}
|
|
}
|
|
|
|
if (*info != 0) {
|
|
i__1 = -(*info);
|
|
xerbla_("DSTEDC", &i__1, (ftnlen)6);
|
|
return;
|
|
} else if (lquery) {
|
|
return;
|
|
}
|
|
|
|
/* Quick return if possible */
|
|
|
|
if (*n == 0) {
|
|
return;
|
|
}
|
|
if (*n == 1) {
|
|
if (icompz != 0) {
|
|
z__[z_dim1 + 1] = 1.;
|
|
}
|
|
return;
|
|
}
|
|
|
|
/* If the following conditional clause is removed, then the routine */
|
|
/* will use the Divide and Conquer routine to compute only the */
|
|
/* eigenvalues, which requires (3N + 3N**2) real workspace and */
|
|
/* (2 + 5N + 2N lg(N)) integer workspace. */
|
|
/* Since on many architectures DSTERF is much faster than any other */
|
|
/* algorithm for finding eigenvalues only, it is used here */
|
|
/* as the default. If the conditional clause is removed, then */
|
|
/* information on the size of workspace needs to be changed. */
|
|
|
|
/* If COMPZ = 'N', use DSTERF to compute the eigenvalues. */
|
|
|
|
if (icompz == 0) {
|
|
dsterf_(n, &d__[1], &e[1], info);
|
|
goto L50;
|
|
}
|
|
|
|
/* If N is smaller than the minimum divide size (SMLSIZ+1), then */
|
|
/* solve the problem with another solver. */
|
|
|
|
if (*n <= smlsiz) {
|
|
|
|
dsteqr_(compz, n, &d__[1], &e[1], &z__[z_offset], ldz, &work[1], info);
|
|
|
|
} else {
|
|
|
|
/* If COMPZ = 'V', the Z matrix must be stored elsewhere for later */
|
|
/* use. */
|
|
|
|
if (icompz == 1) {
|
|
storez = *n * *n + 1;
|
|
} else {
|
|
storez = 1;
|
|
}
|
|
|
|
if (icompz == 2) {
|
|
dlaset_("Full", n, n, &c_b17, &c_b18, &z__[z_offset], ldz);
|
|
}
|
|
|
|
/* Scale. */
|
|
|
|
orgnrm = dlanst_("M", n, &d__[1], &e[1]);
|
|
if (orgnrm == 0.) {
|
|
goto L50;
|
|
}
|
|
|
|
eps = dlamch_("Epsilon");
|
|
|
|
start = 1;
|
|
|
|
/* while ( START <= N ) */
|
|
|
|
L10:
|
|
if (start <= *n) {
|
|
|
|
/* Let FINISH be the position of the next subdiagonal entry */
|
|
/* such that E( FINISH ) <= TINY or FINISH = N if no such */
|
|
/* subdiagonal exists. The matrix identified by the elements */
|
|
/* between START and FINISH constitutes an independent */
|
|
/* sub-problem. */
|
|
|
|
finish = start;
|
|
L20:
|
|
if (finish < *n) {
|
|
tiny = eps * sqrt((d__1 = d__[finish], abs(d__1))) * sqrt((
|
|
d__2 = d__[finish + 1], abs(d__2)));
|
|
if ((d__1 = e[finish], abs(d__1)) > tiny) {
|
|
++finish;
|
|
goto L20;
|
|
}
|
|
}
|
|
|
|
/* (Sub) Problem determined. Compute its size and solve it. */
|
|
|
|
m = finish - start + 1;
|
|
if (m == 1) {
|
|
start = finish + 1;
|
|
goto L10;
|
|
}
|
|
if (m > smlsiz) {
|
|
|
|
/* Scale. */
|
|
|
|
orgnrm = dlanst_("M", &m, &d__[start], &e[start]);
|
|
dlascl_("G", &c__0, &c__0, &orgnrm, &c_b18, &m, &c__1, &d__[
|
|
start], &m, info);
|
|
i__1 = m - 1;
|
|
i__2 = m - 1;
|
|
dlascl_("G", &c__0, &c__0, &orgnrm, &c_b18, &i__1, &c__1, &e[
|
|
start], &i__2, info);
|
|
|
|
if (icompz == 1) {
|
|
strtrw = 1;
|
|
} else {
|
|
strtrw = start;
|
|
}
|
|
dlaed0_(&icompz, n, &m, &d__[start], &e[start], &z__[strtrw +
|
|
start * z_dim1], ldz, &work[1], n, &work[storez], &
|
|
iwork[1], info);
|
|
if (*info != 0) {
|
|
*info = (*info / (m + 1) + start - 1) * (*n + 1) + *info %
|
|
(m + 1) + start - 1;
|
|
goto L50;
|
|
}
|
|
|
|
/* Scale back. */
|
|
|
|
dlascl_("G", &c__0, &c__0, &c_b18, &orgnrm, &m, &c__1, &d__[
|
|
start], &m, info);
|
|
|
|
} else {
|
|
if (icompz == 1) {
|
|
|
|
/* Since QR won't update a Z matrix which is larger than */
|
|
/* the length of D, we must solve the sub-problem in a */
|
|
/* workspace and then multiply back into Z. */
|
|
|
|
dsteqr_("I", &m, &d__[start], &e[start], &work[1], &m, &
|
|
work[m * m + 1], info);
|
|
dlacpy_("A", n, &m, &z__[start * z_dim1 + 1], ldz, &work[
|
|
storez], n);
|
|
dgemm_("N", "N", n, &m, &m, &c_b18, &work[storez], n, &
|
|
work[1], &m, &c_b17, &z__[start * z_dim1 + 1],
|
|
ldz);
|
|
} else if (icompz == 2) {
|
|
dsteqr_("I", &m, &d__[start], &e[start], &z__[start +
|
|
start * z_dim1], ldz, &work[1], info);
|
|
} else {
|
|
dsterf_(&m, &d__[start], &e[start], info);
|
|
}
|
|
if (*info != 0) {
|
|
*info = start * (*n + 1) + finish;
|
|
goto L50;
|
|
}
|
|
}
|
|
|
|
start = finish + 1;
|
|
goto L10;
|
|
}
|
|
|
|
/* endwhile */
|
|
|
|
if (icompz == 0) {
|
|
|
|
/* Use Quick Sort */
|
|
|
|
dlasrt_("I", n, &d__[1], info);
|
|
|
|
} else {
|
|
|
|
/* Use Selection Sort to minimize swaps of eigenvectors */
|
|
|
|
i__1 = *n;
|
|
for (ii = 2; ii <= i__1; ++ii) {
|
|
i__ = ii - 1;
|
|
k = i__;
|
|
p = d__[i__];
|
|
i__2 = *n;
|
|
for (j = ii; j <= i__2; ++j) {
|
|
if (d__[j] < p) {
|
|
k = j;
|
|
p = d__[j];
|
|
}
|
|
/* L30: */
|
|
}
|
|
if (k != i__) {
|
|
d__[k] = d__[i__];
|
|
d__[i__] = p;
|
|
dswap_(n, &z__[i__ * z_dim1 + 1], &c__1, &z__[k * z_dim1
|
|
+ 1], &c__1);
|
|
}
|
|
/* L40: */
|
|
}
|
|
}
|
|
}
|
|
|
|
L50:
|
|
work[1] = (doublereal) lwmin;
|
|
iwork[1] = liwmin;
|
|
|
|
return;
|
|
|
|
/* End of DSTEDC */
|
|
|
|
} /* dstedc_ */
|
|
|