johna@van-bc.wimsey.bc.ca (John Altstadt) (06/15/90)
About a month ago, I posted a request that was met with underwhelming
response. I am assuming that the reason for the lack of response was
because of the lack of source code.
In a nutshell, I wondered if MSC 6.0 had better floating point accuracy
than 5.x.
After some digging, I have been able to find the source code that was
referenced, and I have appended it in a shar, along with my original
posting.
Flames about posting sources to csipp will be graciously ignored.
John
#! /bin/sh
# This is a shell archive, meaning:
# 1. Remove everything above the #! /bin/sh line.
# 2. Save the resulting text in a file.
# 3. Execute the file with /bin/sh (not csh) to create the files:
# msc_vs.tc
# xmemcof.c
# xspctrm.c
# spctrl.dat
# This archive created: Thu Jun 14 12:05:48 1990
export PATH; PATH=/bin:$PATH
echo shar: extracting "'msc_vs.tc'" '(3306 characters)'
if test -f 'msc_vs.tc'
then
echo shar: will not over-write existing file "'msc_vs.tc'"
else
cat << \SHAR_EOF > 'msc_vs.tc'
Now that 6.0 is out, could somebody with the relevant source code
handy (I am missing one of the two books referenced) try out the
"benchmark" below and supply new results for Microsoft? More
than one person must be interested.
Letter from the June, 1989 issue of Byte magazine, pages 34 & 38:
****
Recipe for Comparing Compilers
"Smoothing Out C" by Steve Apiki and Jon Udell (Product Focus,
February) was an informative article. However, the result that
"Turbo... did poorly on the floating-point test" prompts me to
relate data comparing Microsoft C and Turbo C using routines
from the book _Numerical Recipes in C_ by William H. Press et al.
(New York: Cambridge University Press, 1986.) (The
FORTRAN/Pascal version of this book was reviewed in the January
1987 BYTE.)
I needed a microcomputer C compiler for a spectral estimation
project. Originally, I had intended to use the Microsoft
compiler, since several coworkers had been using it and it had
always been reliable. For this project, good numerical accuracy
was far more important than execution speed. Although Microsoft
C and Turbo C produced identical results for most of the
_Numerical Recipes_ routines that I intended to use, Turbo C
provided better accuracy in the two routines memcof.c and
spctrm.c.
I compared compilers by using the driver routines supplied in the
companion book _Numerical Recipes Example Book (C)_. This book
often gives the answers obtained when the routines were run by
the authors. I tested the book's sample routines xspctrm.c and
xmemcof.c using Microsoft C, Turbo C, and VAX-C (the latter on a
DEC VAX-11/780 minicomputer). The results in tables 1 and 2 show
why I selected Turbo C.
Table 1: Results using the xmemcof.c driver routine. Published
results appeared in _Numerical Recipes in C_.
Index Published VAX-C Turbo C Microsoft C
1 1.261539 1.261540 1.261540 1.262617
2 -0.007695 -0.007694 -0.007694 -0.009724
3 -0.646778 -0.646779 -0.646779 -0.646527
4 -0.280603 -0.280603 -0.280603 -0.279950
5 0.163693 0.163692 0.163692 0.163249
6 0.347674 0.347676 0.347676 0.347373
7 0.111247 0.111247 0.111247 0.111045
8 -0.337141 -0.337142 -0.337142 -0.336187
9 -0.358043 -0.358043 -0.358043 -0.358086
10 0.378774 0.378774 0.378774 -0.378403
Table 2: Results using the xspctrm.c driver routine (overlapped
case).
Index Published VAX-C Turbo C Microsoft C
1 - 0.001993 0.001993 0.001786
2 - 0.001461 0.001461 0.002544
3 - 0.072356 0.072356 0.147874
4 - 0.062723 0.062723 0.107431
5 - 0.097625 0.097625 0.000332
6 - 0.014102 0.014102 0.000512
7 - 0.000089 0.000089 0.000113
8 - 0.000210 0.000210 0.000131
9 - 0.000052 0.000052 0.000037
****
I have tried to enter the numbers in the tables as carefully as
possible. I may have messed up a digit or two.
SHAR_EOF
if test 3306 -ne "`wc -c < 'msc_vs.tc'`"
then
echo shar: error transmitting "'msc_vs.tc'" '(should have been 3306 characters)'
fi
fi # end of overwriting check
echo shar: extracting "'xmemcof.c'" '(2203 characters)'
if test -f 'xmemcof.c'
then
echo shar: will not over-write existing file "'xmemcof.c'"
else
cat << \SHAR_EOF > 'xmemcof.c'
/* All required code has been appended to this file -- jma */
/* Driver for routine MEMCOF */
#include <stdio.h>
#include <math.h>
#include <alloc.h>
#include <process.h> /* for exit() prototype */
float *vector(int nl, int nh);
void nrerror(char error_text[]);
void memcof(float data[], int n, int m, float *pm, float cof[]);
void free_vector(float *v, int nl);
#define N 1000
#define M 10
void main(void)
{
int i;
float pm,*cof,*data;
FILE *fp;
cof=vector(1,M);
data=vector(1,N);
if ((fp = fopen("spctrl.dat","r")) == NULL)
nrerror("Data file SPCTRL.DAT not found\n");
for (i=1;i<=N;i++) fscanf(fp,"%f",&data[i]);
fclose(fp);
memcof(data,N,M,&pm,cof);
printf("Coefficients for spectral estimation of SPCTRL.DAT\n\n");
for (i=1;i<=M;i++) printf("a[%2d] = %12.6f \n",i,cof[i]);
printf("\nb0 =%12.6f\n",pm);
free_vector(data,1);
free_vector(cof,1);
}
float *vector(int nl, int nh)
{
float *v;
v=(float *)malloc((unsigned) (nh-nl+1)*sizeof(float));
if (!v) nrerror("allocation failure in vector()");
return v-nl;
}
void nrerror(char error_text[])
{
fprintf(stderr,"Numerical Recipes run-time error...\n");
fprintf(stderr,"%s\n",error_text);
fprintf(stderr,"...now exiting to system...\n");
exit(1);
}
static float sqrarg;
#define SQR(a) (sqrarg=(a),sqrarg*sqrarg)
void memcof(float data[], int n, int m, float *pm, float cof[])
{
int k,j,i;
float p=0.0,*wk1,*wk2,*wkm;
wk1=vector(1,n);
wk2=vector(1,n);
wkm=vector(1,m);
for (j=1;j<=n;j++) p += SQR(data[j]);
*pm=p/n;
wk1[1]=data[1];
wk2[n-1]=data[n];
for (j=2;j<=n-1;j++) {
wk1[j]=data[j];
wk2[j-1]=data[j];
}
for (k=1;k<=m;k++) {
float num=0.0,denom=0.0;
for (j=1;j<=(n-k);j++) {
num += wk1[j]*wk2[j];
denom += SQR(wk1[j])+SQR(wk2[j]);
}
cof[k]=2.0*num/denom;
*pm *= (1.0-SQR(cof[k]));
if (k != 1)
for (i=1;i<=(k-1);i++)
cof[i]=wkm[i]-cof[k]*wkm[k-i];
if (k == m) {
free_vector(wkm,1);
free_vector(wk2,1);
free_vector(wk1,1);
return;
}
for (i=1;i<=k;i++) wkm[i]=cof[i];
for (j=1;j<=(n-k-1);j++) {
wk1[j] -= wkm[k]*wk2[j];
wk2[j]=wk2[j+1]-wkm[k]*wk1[j+1];
}
}
}
#undef SQR
void free_vector(float *v, int nl)
{
free((char*) (v+nl));
}
SHAR_EOF
if test 2203 -ne "`wc -c < 'xmemcof.c'`"
then
echo shar: error transmitting "'xmemcof.c'" '(should have been 2203 characters)'
fi
fi # end of overwriting check
echo shar: extracting "'xspctrm.c'" '(3696 characters)'
if test -f 'xspctrm.c'
then
echo shar: will not over-write existing file "'xspctrm.c'"
else
cat << \SHAR_EOF > 'xspctrm.c'
/* All required code has been appended to this file -- jma */
/* Driver for routine SPCTRM */
#include <stdio.h>
#include <alloc.h>
#include <math.h>
#include <process.h> /* prototype for exit() */
float *vector(int nl, int nh);
void nrerror(char error_text[]);
void spctrm(FILE *fp, float p[], int m, int k, int ovrlap);
void free_vector(float *v, int nl);
void four1(float data[], int nn, int isign);
#define M 16
#define TRUE 1
#define FALSE 0
void main(void)
{
int j,k,ovrlap;
float *p,*q;
FILE *fp;
p=vector(1,M);
q=vector(1,M);
if ((fp = fopen("spctrl.dat","r")) == NULL)
nrerror("Data file SPCTRL.DAT not found\n");
k=8;
ovrlap=TRUE;
spctrm(fp,p,M,k,ovrlap);
rewind(fp);
k=16;
ovrlap=FALSE;
spctrm(fp,q,M,k,ovrlap);
fclose(fp);
printf("\nSpectrum of data in file SPCTRL.DAT\n");
printf("%13s %s %5s %s\n"," ","overlapped "," ","non-overlapped");
for (j=1;j<=M;j++)
printf("%3d %5s %13f %5s %13f\n",j," ",p[j]," ",q[j]);
free_vector(q,1);
free_vector(p,1);
}
float *vector(int nl, int nh)
{
float *v;
v=(float *)malloc((unsigned) (nh-nl+1)*sizeof(float));
if (!v) nrerror("allocation failure in vector()");
return v-nl;
}
void nrerror(char error_text[])
{
fprintf(stderr,"Numerical Recipes run-time error...\n");
fprintf(stderr,"%s\n",error_text);
fprintf(stderr,"...now exiting to system...\n");
exit(1);
}
static float sqrarg;
#define SQR(a) (sqrarg=(a),sqrarg*sqrarg)
#define WINDOW(j,a,b) (1.0-fabs((((j)-1)-(a))*(b))) /* Parzen */
/* #define WINDOW(j,a,b) 1.0 */ /* Square */
/* #define WINDOW(j,a,b) (1.0-SQR((((j)-1)-(a))*(b))) */ /* Welch */
void spctrm(FILE *fp, float p[], int m, int k, int ovrlap)
{
int mm,m44,m43,m4,kk,joffn,joff,j2,j;
float w,facp,facm,*w1,*w2,sumw=0.0,den=0.0;
mm=m+m;
m43=(m4=mm+mm)+3;
m44=m43+1;
w1=vector(1,m4);
w2=vector(1,m);
facm=m-0.5;
facp=1.0/(m+0.5);
for (j=1;j<=mm;j++) sumw += SQR(WINDOW(j,facm,facp));
for (j=1;j<=m;j++) p[j]=0.0;
if (ovrlap)
for (j=1;j<=m;j++) fscanf(fp,"%f",&w2[j]);
for (kk=1;kk<=k;kk++) {
for (joff = -1;joff<=0;joff++) {
if (ovrlap) {
for (j=1;j<=m;j++) w1[joff+j+j]=w2[j];
for (j=1;j<=m;j++) fscanf(fp,"%f",&w2[j]);
joffn=joff+mm;
for (j=1;j<=m;j++) w1[joffn+j+j]=w2[j];
} else {
for (j=joff+2;j<=m4;j+=2)
fscanf(fp,"%f",&w1[j]);
}
}
for (j=1;j<=mm;j++) {
j2=j+j;
w=WINDOW(j,facm,facp);
w1[j2] *= w;
w1[j2-1] *= w;
}
four1(w1,mm,1);
p[1] += (SQR(w1[1])+SQR(w1[2]));
for (j=2;j<=m;j++) {
j2=j+j;
p[j] += (SQR(w1[j2])+SQR(w1[j2-1])
+SQR(w1[m44-j2])+SQR(w1[m43-j2]));
}
den += sumw;
}
den *= m4;
for (j=1;j<=m;j++) p[j] /= den;
free_vector(w2,1);
free_vector(w1,1);
}
#undef SQR
#undef WINDOW
void free_vector(float *v, int nl)
{
free((char*) (v+nl));
}
#define SWAP(a,b) tempr=(a);(a)=(b);(b)=tempr
void four1(float data[], int nn, int isign)
{
int n,mmax,m,j,istep,i;
double wtemp,wr,wpr,wpi,wi,theta;
float tempr,tempi;
n=nn << 1;
j=1;
for (i=1;i<n;i+=2) {
if (j > i) {
SWAP(data[j],data[i]);
SWAP(data[j+1],data[i+1]);
}
m=n >> 1;
while (m >= 2 && j > m) {
j -= m;
m >>= 1;
}
j += m;
}
mmax=2;
while (n > mmax) {
istep=2*mmax;
theta=6.28318530717959/(isign*mmax);
wtemp=sin(0.5*theta);
wpr = -2.0*wtemp*wtemp;
wpi=sin(theta);
wr=1.0;
wi=0.0;
for (m=1;m<mmax;m+=2) {
for (i=m;i<=n;i+=istep) {
j=i+mmax;
tempr=wr*data[j]-wi*data[j+1];
tempi=wr*data[j+1]+wi*data[j];
data[j]=data[i]-tempr;
data[j+1]=data[i+1]-tempi;
data[i] += tempr;
data[i+1] += tempi;
}
wr=(wtemp=wr)*wpr-wi*wpi+wr;
wi=wi*wpr+wtemp*wpi+wi;
}
mmax=istep;
}
}
#undef SWAP
SHAR_EOF
if test 3696 -ne "`wc -c < 'xspctrm.c'`"
then
echo shar: error transmitting "'xspctrm.c'" '(should have been 3696 characters)'
fi
fi # end of overwriting check
echo shar: extracting "'spctrl.dat'" '(15000 characters)'
if test -f 'spctrl.dat'
then
echo shar: will not over-write existing file "'spctrl.dat'"
else
cat << \SHAR_EOF > 'spctrl.dat'
0.097911 0.367270 0.500363 0.425597
0.058794 -0.383898 -0.811678 -0.852240
-0.540416 0.077375 0.609527 0.949625
0.934536 0.547584 0.065175 -0.343537
-0.364521 -0.234990 -0.060539 0.070118
-0.053797 -0.195746 -0.378981 -0.302282
0.078039 0.496074 0.873213 0.920654
0.639311 0.016809 -0.545344 -0.847283
-0.757841 -0.419941 0.050494 0.412930
0.443676 0.355997 0.190416 0.004934
0.098812 0.334695 0.487672 0.390609
0.091482 -0.471872 -0.822832 -0.862922
-0.541732 0.037119 0.632220 0.969027
0.938331 0.524281 0.054555 -0.316386
-0.417168 -0.258567 -0.002089 0.081358
-0.010683 -0.250513 -0.395420 -0.292397
0.037531 0.541725 0.858324 0.979284
0.632282 0.052913 -0.513068 -0.861395
-0.844621 -0.381213 0.032628 0.441947
0.494978 0.353964 0.103450 0.050774
0.186340 0.348806 0.477618 0.439581
0.030414 -0.376526 -0.748378 -0.839505
-0.572722 0.070455 0.592158 0.997204
0.871991 0.573119 0.016951 -0.299401
-0.380699 -0.213378 -0.045667 0.014124
-0.083163 -0.211765 -0.394188 -0.329720
0.056738 0.487655 0.899056 0.923687
0.679856 0.053336 -0.580099 -0.898054
-0.777465 -0.449933 0.008536 0.397103
0.442733 0.382489 0.126694 0.008359
0.174158 0.298124 0.492837 0.387331
0.092230 -0.421991 -0.760395 -0.836616
-0.490595 0.002870 0.662405 0.970864
0.870277 0.485316 0.037991 -0.338729
-0.430330 -0.232343 -0.048044 0.032509
-0.019183 -0.230955 -0.379461 -0.286177
0.009042 0.568075 0.874133 0.912020
0.610105 0.098793 -0.487907 -0.815460
-0.839317 -0.455718 0.033628 0.359049
0.489247 0.299280 0.142992 0.076076
0.164557 0.331189 0.435143 0.371382
0.028573 -0.429901 -0.789573 -0.827289
-0.516383 0.019417 0.660162 0.998339
0.943339 0.553927 0.064096 -0.268127
-0.343639 -0.208262 -0.080817 0.014051
-0.091364 -0.251841 -0.332385 -0.253401
0.004142 0.562365 0.857352 0.980905
0.640451 0.070460 -0.560346 -0.860592
-0.800160 -0.411101 0.035681 0.419918
0.456588 0.380944 0.103726 0.061474
0.146002 0.335740 0.530424 0.424361
0.001236 -0.376993 -0.768314 -0.816661
-0.574880 0.001497 0.631647 0.908096
0.882109 0.558594 0.042774 -0.296324
-0.405426 -0.211205 -0.029234 0.085678
-0.036877 -0.276193 -0.370418 -0.333459
0.030051 0.509729 0.935126 0.962784
0.662392 0.072465 -0.579946 -0.894496
-0.846667 -0.445903 0.094651 0.436459
0.451768 0.372502 0.171457 0.074792
0.140627 0.311938 0.514634 0.414529
0.037810 -0.377700 -0.813998 -0.880491
-0.504330 0.016679 0.596882 0.918905
0.889236 0.478435 0.036437 -0.323710
-0.398283 -0.293001 -0.081568 0.036554
-0.019824 -0.221855 -0.387019 -0.334675
0.080294 0.491516 0.922518 0.963173
0.661262 0.013510 -0.562322 -0.893592
-0.764120 -0.466779 0.016865 0.438673
0.475335 0.347867 0.122441 0.095509
0.122242 0.374897 0.432223 0.394942
0.089702 -0.404219 -0.756637 -0.845088
-0.516642 0.039524 0.654199 0.984378
0.896149 0.525603 0.033058 -0.247243
-0.355035 -0.228370 0.007792 0.008678
-0.064855 -0.231288 -0.374441 -0.332992
0.005931 0.508226 0.942214 1.002574
0.586029 0.075539 -0.508032 -0.893993
-0.798908 -0.387166 0.083562 0.384966
0.487865 0.332096 0.181928 0.008358
0.135767 0.352306 0.478585 0.401968
0.054226 -0.399437 -0.819389 -0.826667
-0.533762 0.002544 0.677559 0.921943
0.884705 0.488659 0.032441 -0.260313
-0.387379 -0.218315 -0.088176 0.045573
-0.084575 -0.283227 -0.431495 -0.305018
0.020692 0.495659 0.903865 0.979479
0.641578 0.096499 -0.532013 -0.818501
-0.801651 -0.458958 0.051335 0.414244
0.489763 0.336589 0.117080 0.084142
0.125365 0.301872 0.512340 0.445291
0.073475 -0.448532 -0.801094 -0.842954
-0.528140 0.025966 0.603184 0.946022
0.882719 0.539963 0.098792 -0.289700
-0.387006 -0.195881 -0.081170 0.023650
-0.026617 -0.253341 -0.408916 -0.343958
0.066436 0.554518 0.915065 0.979764
0.615021 0.095966 -0.535727 -0.900540
-0.747872 -0.456808 0.018016 0.432158
0.453624 0.354854 0.139914 0.073312
0.107453 0.362039 0.483963 0.428043
0.095897 -0.416062 -0.752085 -0.881777
-0.506618 0.019157 0.634220 0.906013
0.942182 0.536385 0.055480 -0.316476
-0.351742 -0.213961 -0.053189 0.060703
0.001519 -0.246973 -0.384560 -0.331215
0.009580 0.493835 0.911937 0.951883
0.603137 0.054710 -0.524431 -0.897798
-0.780356 -0.408528 0.033938 0.424251
0.528502 0.376326 0.176091 0.045445
0.164536 0.375618 0.525160 0.426136
0.008985 -0.447627 -0.807413 -0.859091
-0.559236 0.011335 0.590186 0.936487
0.856904 0.530043 0.012522 -0.335937
-0.412646 -0.209519 -0.053693 0.051147
-0.053526 -0.222614 -0.410404 -0.279051
0.019434 0.520151 0.924055 0.905338
0.599860 0.072868 -0.495636 -0.852138
-0.842475 -0.410295 0.045931 0.374146
0.507044 0.304955 0.110229 0.014329
0.099445 0.313920 0.442175 0.440693
0.057408 -0.424944 -0.804769 -0.813377
-0.567945 0.061968 0.598933 0.986802
0.848188 0.515720 0.047513 -0.329140
-0.349152 -0.242236 -0.078811 0.079854
-0.024022 -0.265492 -0.355148 -0.309431
0.096588 0.575235 0.849259 0.941124
0.595528 0.055573 -0.503313 -0.883673
-0.816723 -0.439862 0.048003 0.440092
0.494922 0.379278 0.180705 0.057666
0.108089 0.332832 0.464212 0.374919
0.021429 -0.471234 -0.757271 -0.809901
-0.497724 0.082059 0.657500 0.948854
0.896929 0.511290 0.073270 -0.317564
-0.361846 -0.280574 -0.033892 0.037408
-0.073339 -0.285403 -0.429735 -0.246847
0.031412 0.567831 0.933083 0.974899
0.648928 0.090599 -0.513837 -0.874769
-0.778362 -0.450210 0.063260 0.403755
0.485675 0.325856 0.126301 0.056894
0.153595 0.366111 0.464019 0.377563
0.004612 -0.414628 -0.761011 -0.867799
-0.499577 0.091861 0.654241 0.997871
0.942204 0.483009 0.091923 -0.335708
-0.338226 -0.278998 -0.008505 0.053219
-0.058173 -0.219469 -0.352340 -0.302103
0.052941 0.528985 0.893284 0.981896
0.606492 0.065509 -0.534164 -0.857461
-0.802981 -0.388117 0.004868 0.442722
0.435563 0.314212 0.183261 0.077975
0.176073 0.295736 0.482821 0.374858
0.076716 -0.422508 -0.837674 -0.807773
-0.531693 0.074174 0.649856 0.953869
0.851023 0.573365 0.006548 -0.324274
-0.335403 -0.199929 0.004564 0.097017
-0.040870 -0.271010 -0.415173 -0.263773
0.018949 0.509171 0.912486 0.909274
0.663260 0.016081 -0.488704 -0.868258
-0.747901 -0.388979 0.024781 0.367971
0.522698 0.305092 0.186602 0.098019
0.166066 0.381537 0.442624 0.424076
0.025891 -0.452799 -0.814807 -0.810631
-0.572332 0.070288 0.605258 0.956582
0.903865 0.495041 0.050691 -0.276583
-0.421594 -0.278223 -0.006641 0.091677
-0.035410 -0.200884 -0.356414 -0.292225
0.088560 0.537159 0.860301 0.960125
0.597692 0.067915 -0.500009 -0.852144
-0.810110 -0.401515 0.002545 0.396196
0.437215 0.334308 0.113528 0.034839
0.154359 0.297152 0.512050 0.402023
0.053790 -0.411957 -0.776809 -0.804607
-0.545230 0.004483 0.606468 1.004181
0.877637 0.520703 0.014247 -0.327739
-0.427206 -0.228288 -0.067409 0.002191
-0.059131 -0.221008 -0.343540 -0.253384
0.047465 0.485443 0.899890 0.910847
0.663669 0.037764 -0.517929 -0.876655
-0.751527 -0.433621 0.018924 0.441145
0.523340 0.391304 0.156069 0.020190
0.184455 0.323097 0.455107 0.361767
0.089370 -0.385349 -0.794025 -0.903242
-0.532589 0.034371 0.668037 0.928333
0.929222 0.547698 0.097208 -0.311597
-0.367236 -0.232687 -0.051130 0.048259
-0.060719 -0.260017 -0.418074 -0.303872
0.072928 0.484907 0.855767 0.976362
0.610166 0.014785 -0.486794 -0.846244
-0.748778 -0.394212 0.009820 0.358356
0.526840 0.354162 0.190473 0.028436
0.176236 0.322849 0.462059 0.370479
0.064294 -0.384773 -0.792332 -0.806280
-0.523412 0.078413 0.661642 0.926860
0.922979 0.544351 0.025883 -0.340230
-0.354519 -0.247419 -0.003119 0.082101
-0.012022 -0.198200 -0.391275 -0.330999
0.072228 0.554047 0.940904 0.963250
0.607121 0.019336 -0.555746 -0.832664
-0.760450 -0.395728 0.061648 0.383536
0.522079 0.316498 0.134781 0.021790
0.168653 0.344387 0.485338 0.429979
0.091185 -0.386099 -0.803865 -0.847632
-0.547286 0.015153 0.586225 0.939693
0.849235 0.559308 0.012852 -0.294433
-0.399566 -0.258405 -0.066580 0.087800
-0.042150 -0.269474 -0.332681 -0.322917
0.056257 0.541328 0.884606 0.941629
0.630189 0.065754 -0.561263 -0.903853
-0.790730 -0.450923 0.096406 0.379047
0.518242 0.362239 0.096277 0.042876
0.161780 0.378011 0.458697 0.358680
0.082284 -0.444665 -0.803149 -0.815407
-0.489638 0.091010 0.635438 0.941832
0.856547 0.564492 0.099813 -0.318179
-0.419672 -0.290083 -0.013989 0.083569
-0.055919 -0.257622 -0.430436 -0.277385
0.006100 0.515979 0.908006 0.993760
0.618310 0.001543 -0.535351 -0.899117
-0.829096 -0.456206 0.000256 0.376039
0.480801 0.382116 0.150624 0.018162
0.166605 0.306360 0.447567 0.406553
0.077535 -0.439301 -0.842485 -0.819257
-0.544317 0.078440 0.623969 0.977282
0.903853 0.526331 0.094650 -0.329740
-0.350568 -0.286092 -0.062614 0.015146
-0.088102 -0.281699 -0.397238 -0.311890
0.036393 0.481938 0.889484 0.955059
0.662317 0.005069 -0.503866 -0.861251
-0.781591 -0.391141 0.042585 0.378318
0.520022 0.321098 0.110956 0.008048
0.139262 0.313787 0.474640 0.407562
0.058446 -0.454213 -0.845370 -0.813975
-0.480999 0.010004 0.665455 0.931345
0.931231 0.562228 0.023416 -0.256999
-0.333008 -0.200222 -0.044808 0.065249
-0.071637 -0.273911 -0.332033 -0.306899
0.020117 0.480633 0.907756 0.986501
0.596889 0.038615 -0.566205 -0.817764
-0.756176 -0.403098 0.035555 0.444095
0.510418 0.307940 0.179051 0.061500
0.098928 0.345776 0.464550 0.435004
0.060468 -0.383262 -0.788087 -0.831224
-0.523793 0.033440 0.633217 0.944808
0.920680 0.483308 0.058446 -0.254890
-0.402245 -0.206168 0.002455 0.098080
0.001098 -0.204013 -0.342597 -0.276006
0.076104 0.519721 0.848375 0.954724
0.595388 0.017591 -0.554098 -0.823085
-0.832739 -0.394980 0.031532 0.407808
0.506202 0.365729 0.112260 0.015105
0.156834 0.390595 0.502499 0.376821
0.022250 -0.468862 -0.828902 -0.865065
-0.521754 0.008702 0.582262 0.990812
0.858668 0.479814 0.004871 -0.339759
-0.382635 -0.254824 -0.058774 0.040762
-0.064244 -0.207980 -0.406088 -0.248284
0.065140 0.502790 0.909789 0.948749
0.594440 0.095172 -0.482355 -0.813955
-0.750694 -0.449014 0.038131 0.432970
0.460675 0.310682 0.140371 0.003175
0.148347 0.304243 0.480985 0.429534
0.059744 -0.451427 -0.761063 -0.903347
-0.542992 0.066439 0.606244 0.972373
0.849850 0.538147 0.037460 -0.270762
-0.385688 -0.243512 -0.051111 0.040176
-0.046009 -0.222313 -0.415061 -0.344722
0.065446 0.521380 0.896401 0.990178
0.672997 0.031288 -0.488687 -0.839813
-0.747661 -0.454405 0.003323 0.431151
0.498745 0.359902 0.147358 0.026276
0.119546 0.357631 0.468306 0.368227
0.007975 -0.405726 -0.835639 -0.826552
-0.498900 0.075505 0.610725 0.999256
0.849490 0.538016 0.030475 -0.274609
-0.394095 -0.244471 -0.007600 0.048182
-0.090882 -0.213751 -0.363397 -0.248384
0.085440 0.574230 0.947322 0.982645
0.653703 0.025142 -0.537558 -0.809595
-0.807243 -0.444492 0.099258 0.365283
0.499182 0.301096 0.155158 0.094121
0.180500 0.367823 0.453490 0.358748
0.090581 -0.409058 -0.779713 -0.805878
-0.572210 0.007475 0.614132 0.921802
0.927101 0.504223 0.068253 -0.268464
-0.380833 -0.238024 -0.012982 0.007172
-0.009599 -0.259873 -0.391673 -0.293916
0.070242 0.564174 0.881125 0.968656
0.648719 0.063631 -0.568027 -0.824792
-0.806184 -0.454851 0.026379 0.404643
0.448427 0.351062 0.126076 0.097715
0.138137 0.355728 0.525115 0.349266
0.055674 -0.447023 -0.779438 -0.871175
-0.572309 0.074596 0.583332 0.926666
0.866427 0.571239 0.061118 -0.282493
-0.340329 -0.231417 -0.031789 0.092093
-0.067833 -0.236263 -0.415730 -0.255997
0.061875 0.484668 0.912390 0.998998
0.608189 0.011819 -0.575031 -0.844429
-0.785034 -0.465118 0.005862 0.381676
0.500147 0.331682 0.186613 0.002278
SHAR_EOF
if test 15000 -ne "`wc -c < 'spctrl.dat'`"
then
echo shar: error transmitting "'spctrl.dat'" '(should have been 15000 characters)'
fi
fi # end of overwriting check
# End of shell archive
exit 0
--
John Altstadt, 6135 Carson St., Burnaby B.C., CANADA, V5J 2Z8
..!ubc-cs!van-bc!johna || ..!uunet!van-bc!johna || johna@wimsey.bc.ca
There may, or there may not have been some :-)s up there. There should
probably have been a few more for the humorless, lost souls out there.kdq@demott.COM (Kevin D. Quitt) (06/16/90)
Results from MSC 6.0
First column under Microsoft uses in-line 80x87 instructions
(/FPi87), second column is for both 8087-compatible library calls (no
fpu: /FPc), and for their alternate math library (/FPa). All
compilations were otherwise: /AS /Od /Zp1 /G1
Table 1: Results using the xmemcof.c driver routine. Published
results appeared in _Numerical Recipes in C_.
Index Published VAX-C Turbo C Microsoft C
1 1.261539 1.261540 1.261540 1.261539 1.261540
2 -0.007695 -0.007694 -0.007694 -0.007695 -0.007693
3 -0.646778 -0.646779 -0.646779 -0.646777 -0.646780
4 -0.280603 -0.280603 -0.280603 -0.280602 -0.280604
5 0.163693 0.163692 0.163692 0.163691 0.163693
6 0.347674 0.347676 0.347676 0.347675 0.347676
7 0.111247 0.111247 0.111247 0.111247 0.111247
8 -0.337141 -0.337142 -0.337142 -0.337141 -0.337142
9 -0.358043 -0.358043 -0.358043 -0.358044 -0.358043
10 0.378774 0.378774 0.378774 0.378775 0.378774
Table 2: Results using the xspctrm.c driver routine (overlapped
case).
Index Published VAX-C Turbo C Microsoft C
1 - 0.001993 0.001993 0.001993 0.001993
2 - 0.001461 0.001461 0.001461 0.001461
3 - 0.072356 0.072356 0.072356 0.072356
4 - 0.062723 0.062723 0.062723 0.062723
5 - 0.097625 0.097625 0.097625 0.097625
6 - 0.014102 0.014102 0.014102 0.014102
7 - 0.000089 0.000089 0.000089 0.000089
8 - 0.000210 0.000210 0.000210 0.000210
9 - 0.000052 0.000052 0.000052 0.000052
Compiling with 5.1, I was able to duplicate the erroneous results
only by selecting the emulation library, and not using the 80x87
instructions. This is a farily bizarre thing to do if you don't have a
coprocessor, you should be using the alt-math library, since it's
significantly faster than the emulator. It also produces the correct
answers. I have yet to find a compiler (on any machine) whose default
switches were satisfactory even most of the time.
--
_
Kevin D. Quitt demott!kdq kdq@demott.com
DeMott Electronics Co. 14707 Keswick St. Van Nuys, CA 91405-1266
VOICE (818) 988-4975 FAX (818) 997-1190 MODEM (818) 997-4496 PEP last
96.37% of all statistics are made up.