Bspline implementation not working as expected - c

I made this implementation of the BSPLINE curve. I Followed the usual definition presented in http://en.wikipedia.org/wiki/B-spline
t is the knot vector.
#include <stdio.h>
double N(int i, int k, double u, double t[])
{
if(k == 1)
{
if(u >= t[i] && u < t[i+1])
return 1.0e0;
else {
return 0.0e0;
}
}
return ((u - t[i])*N(i, k -1, u, t))/(t[i+k-1] - t[i]) + ((t[i+k] - u)*N(i+1, k-1, u, t))/(t[i+k] - t[i+1]);
}
double pu(double u, double x[], int n, int k, double t[])
{
int i;
double r = 0.0e0;
for(i = 0; i < n; i++)
{
r += x[i]*N(i, k, u, t);
}
return r;
}
int main()
{
double t[] = {0.0, 0.5, 1, 2, 3, 4, 4.5, 5}; //knot vector
double x[] = {-30.0, 25.0, 9.0, 20.0, 25.0, 31.0}, y[] = {-5.0, -10.0, 3.0, -10.0, -5.0, 25.0}; //the points
double u;
for(u = 0.0e0; u < 5.0; u+=0.01e0)
{
printf("%lf %lf\n", pu(u, x, 6, 2, t), pu(u, y, 6, 2, t));
}
return 0;
}
The problem is that when I plot the computed points I see that there is an unexpected behavior at the beginning and end of the curve.
For instance:
I can not understand why this happens, try changing the values of t, but it seems that is not it.

There are two implementations of bspline: uniform and standard. In uniform the first and the last control points are not interpolated and in the standard knot sequence both are interpolated. In uniform, you can have uniform knots usually 1,2,3,... For standard knot sequence if you have order k (degree k-1) you should have k zeros, k ones and fill in the middle with 1/(m-k+2) where m is the number of control points. For example, having 5 control points and order 3, knot sequences are 0, 0, 0, 0.25, 0.5, 0.75, 1, 1, 1.
In addition, using delta functions you can have a much better implementation rather than computing N function. Delta function benefits from the local support of bspline. I suggest you take a look at the course notes that I teach at the University of Calgary:
http://pages.cpsc.ucalgary.ca/~amahdavi/pmwiki-2.2.8/uploads/Site/notes1.pdf
Check page 40 algorithm 3.3.
Hopefully it's helpful.

In general: if you have n control points and you construct a bspline curve of degree k, your knot vector has n+k+1 knots. The domain of the curve (i.e. where the summation of the basis functions are =1) is given in the interval [t_k, ... t_n] (as far as my details are right).
In the loop, your parameter values range from 0.0 to 5.0. This should be t[k] = t[2] = 1.0 to t[n] = t[6] = 4.0.
The wiki examples do it the other way, as the knot vectors given there have multiple values (i.e. k-times) at the beginning and the end. So the shown spline curve starts/ends at the first/last control point. Your construction does not give this property.

Related

Initial value problem for a system of ODEs solver C program

So I wanted to implement the path of the Moon around the Earth with a C program.
My problem is that you know the Moon's velocity and position at Apogee and Perigee.
So I started to solve it from Apogee, but I cannot figure out how I could add the second velocity and position as "initial value" for it. I tried it with an if but I don't see any difference between the results. Any help is appreciated!
Here is my code:
#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#include <string.h>
typedef void (*ode)(double* p, double t, double* k, double* dk);
void euler(ode f, double *p, double t, double* k, double h, int n, int N)
{
double kn[N];
double dk[N];
double Rp = - 3.633 * pow(10,8); // x position at Perigee
for(int i = 0; i < n; i++)
{
f(p, 0, k, dk);
for (int j = 0; j < N; j++)
{
if (k[0] == Rp) // this is the "if" I mentioned in my comment
// x coordinate at Perigee
{
k[1] = 0; // y coordinate at Perigee
k[2] = 0; // x velocity component at Perigee
k[3] = 1076; // y velocity component at Perigee
}
kn[j] = k[j] + h * dk[j];
printf("%f ", kn[j]);
k[j] = kn[j];
}
printf("\n");
}
}
void gravity_equation(double* p, double t, double* k, double* dk)
{
// Earth is at the (0, 0)
double G = p[0]; // Gravitational constant
double m = p[1]; // Earth mass
double x = k[0]; // x coordinate at Apogee
double y = k[1]; // y coordinate at Apogee
double Vx = k[2]; // x velocity component at Apogee
double Vy = k[3]; // y velocity component at Apogee
dk[0] = Vx;
dk[1] = Vy;
dk[2] = (- G * m * x) / pow(sqrt((x * x)+(y * y)),3);
dk[3] = (- G * m * y) / pow(sqrt((x * x)+(y * y)),3);
}
void run_gravity_equation()
{
int N = 4; // how many equations there are
double initial_values[N];
initial_values[0] = 4.055*pow(10,8); // x position at Apogee
initial_values[1] = 0; // y position at Apogee
initial_values[2] = 0; // x velocity component at Apogee
initial_values[3] = (-1) * 964; //y velocity component at Perigee
int p = 2; // how many parameters there are
double parameters[p];
parameters[0] = 6.67384 * pow(10, -11); // Gravitational constant
parameters[1] = 5.9736 * pow(10, 24); // Earth mass
double h = 3600; // step size
int n = 3000; // the number of steps
euler(&gravity_equation, parameters, 0, initial_values, h, n, N);
}
int main()
{
run_gravity_equation();
return 0;
}
Your interface is
euler(odefun, params, t0, y0, h, n, N)
where
N = dimension of state space
n = number of steps to perform
h = step size
t0, y0 = initial time and value
The intended function of this procedure seems to be that the updated values are returned inside the array y0. There is no reason to insert some hack to force the state to have some initial conditions. The initial condition is passed as argument. As you are doing in void run_gravity_equation(). The integration routine should remain agnostic of the details of the physical model.
It is extremely improbable that you will hit the same value in k[0] == Rp a second time. What you can do is to check for sign changes in Vx, that is, k[1] to find points or segments of extremal x coordinate.
Trying to interpret your description closer, what you want to do is to solve a boundary value problem where x(0)=4.055e8, x'(0)=0, y'(0)=-964 and x(T)=-3.633e8, x'(T)=0. This has the advanced tasks to solve a boundary value problem with single or multiple shooting and additionally, that the upper boundary is variable.
You might want to to use the Kepler laws to get further insights into the parameters of this problem so that you can solve it just with a forward integration. The Kepler ellipse of the first Kepler law has the formula (scaled for Apogee at phi=0, Perigee at phi=pi)
r = R/(1-E*cos(phi))
so that
R/(1-E)=4.055e8 and R/(1+E)=3.633e8,
which gives
R=3.633*(1+E)=4.055*(1-E)
==> E = (4.055-3.633)/(4.055+3.633) = 0.054891,
R = 3.633e8*(1+0.05489) = 3.8324e8
Further, the angular velocity is given by the second Kepler law
phi'*r^2 = const. = sqrt(R*G*m)
which gives tangential velocities at Apogee (r=R/(1-E))
y'(0)=phi'*r = sqrt(R*G*m)*(1-E)/R = 963.9438
and Perigee (r=R/(1+E))
-y'(T)=phi'*r = sqrt(R*G*m)*(1+E)/R = 1075.9130
which indeed reproduces the constants you used in your code.
The area of the Kepler ellipse is pi/4 times the product of smallest and largest diameter. The smallest diameter can be found at cos(phi)=E, the largest is the sum of apogee and perigee radius, so that the area is
pi*R/sqrt(1-E^2)*(R/(1+E)+R/(1-E))/2= pi*R^2/(1-E^2)^1.5
At the same time it is the integral over 0.5*phi*r^2 over the full period 2*T, thus equal to
sqrt(R*G*m)*T
which is the third Kepler law. This allows to compute the half-period as
T = pi/sqrt(G*m)*(R/(1-E^2))^1.5 = 1185821
With h = 3600 the half point should be reached between n=329 and n=330 (n=329.395). Integration with scipy.integrate.odeint vs. Euler steps gives the following table for h=3600:
n [ x[n], y[n] ] for odeint/lsode for Euler
328 [ -4.05469444e+08, 4.83941626e+06] [ -4.28090166e+08, 3.81898023e+07]
329 [ -4.05497554e+08, 1.36933874e+06] [ -4.28507841e+08, 3.48454695e+07]
330 [ -4.05494242e+08, -2.10084488e+06] [ -4.28897657e+08, 3.14986514e+07]
The same for h=36, n=32939..32940
n [ x[n], y[n] ] for odeint/lsode for Euler
32938 [ -4.05499997e+08 5.06668940e+04] [ -4.05754415e+08 3.93845978e+05]
32939 [ -4.05500000e+08 1.59649309e+04] [ -4.05754462e+08 3.59155385e+05]
32940 [ -4.05500000e+08 -1.87370323e+04] [ -4.05754505e+08 3.24464789e+05]
32941 [ -4.05499996e+08 -5.34389954e+04] [ -4.05754545e+08 2.89774191e+05]
which is a little closer for the Euler method, but not much better.

Eigenvalue calculation using TQLI algorithm fails with segmentation fault

I am trying to calculate eigenvalues using the TQLI algorithm that I got from the website of the CACS of the University of Southern California. My test script looks like this:
#include <stdio.h>
int main()
{
int i;
i = rand();
printf("My random number: %d\n", i);
float d[4] = {
{1, 2, 3, 4}
};
float e[4] = {
{0, 0, 0, 0}
};
float z[4][4] = {
{1.0, 0.0, 0.0, 0.0} ,
{0.0, 1.0, 0.0, 0.0} ,
{0.0, 0.0, 1.0, 0.0},
{0.0, 0.0, 0.0, 1.0}
};
double *zptr;
zptr = &z[0][0];
printf("Element [2][1] of identity matrix: %f\n", z[2][1]);
printf("Element [2][2] of identity matrix: %f\n", z[2][2]);
tqli(d, e, 4, zptr);
printf("First eigenvalue: %f\n", d[0]);
return 0;
}
When I try to run this script I get a segmentation fault error as you can see in here. At what location does my code produce this segmentation fault. As I believe the code from USC is bug-free I am pretty sure the mistake must be in my call of the function. However I can't see where I made a mistake in my set-up of the arrays as in my opinion I followed the instructions.
Eigenvalue calculation using TQLI algorithm fails with segmentation
fault
Segmentation fault comes from crossing the supplied array boundary. tqli requires specific data preparation.
1) The eigen code from CACS is Fortran based and counts indexes from 1.
2) The tqli expects double pointer for its matrix and double vectors.
/******************************************************************************/
void tqli(double d[], double e[], int n, double **z)
/*******************************************************************************
d, and e should be declared as double.
3) The program needs modification in respect to the data preparation for the above function.
Helper 1-index based vectors have to be created to supply properly formatted data for the tqli:
double z[NP][NP] = { {2, 0, 0}, {0, 4, 0}, {0, 0, 2} } ;
double **a;
double *d,*e,*f;
d=dvector(1,NP); // 1-index based vector
e=dvector(1,NP);
f=dvector(1,NP);
a=dmatrix(1,NP,1,NP); // 1-index based matrix
for (i=1;i<=NP;i++) // loading data from zero besed `ze` to `a`
for (j=1;j<=NP;j++) a[i][j]=z[i-1][j-1];
Complete test program is supplied below. It uses the eigen code from CACS:
/*******************************************************************************
Eigenvalue solvers, tred2 and tqli, from "Numerical Recipes in C" (Cambridge
Univ. Press) by W.H. Press, S.A. Teukolsky, W.T. Vetterling, and B.P. Flannery
*******************************************************************************/
#include <stdio.h>
#include <stdlib.h>
#include <math.h>
#define NR_END 1
#define SIGN(a,b) ((b) >= 0.0 ? fabs(a) : -fabs(a))
double **dmatrix(int nrl, int nrh, int ncl, int nch)
/* allocate a double matrix with subscript range m[nrl..nrh][ncl..nch] */
{
int i,nrow=nrh-nrl+1,ncol=nch-ncl+1;
double **m;
/* allocate pointers to rows */
m=(double **) malloc((size_t)((nrow+NR_END)*sizeof(double*)));
m += NR_END;
m -= nrl;
/* allocate rows and set pointers to them */
m[nrl]=(double *) malloc((size_t)((nrow*ncol+NR_END)*sizeof(double)));
m[nrl] += NR_END;
m[nrl] -= ncl;
for(i=nrl+1;i<=nrh;i++) m[i]=m[i-1]+ncol;
/* return pointer to array of pointers to rows */
return m;
}
double *dvector(int nl, int nh)
/* allocate a double vector with subscript range v[nl..nh] */
{
double *v;
v=(double *)malloc((size_t) ((nh-nl+1+NR_END)*sizeof(double)));
return v-nl+NR_END;
}
/******************************************************************************/
void tred2(double **a, int n, double d[], double e[])
/*******************************************************************************
Householder reduction of a real, symmetric matrix a[1..n][1..n].
On output, a is replaced by the orthogonal matrix Q effecting the
transformation. d[1..n] returns the diagonal elements of the tridiagonal matrix,
and e[1..n] the off-diagonal elements, with e[1]=0. Several statements, as noted
in comments, can be omitted if only eigenvalues are to be found, in which case a
contains no useful information on output. Otherwise they are to be included.
*******************************************************************************/
{
int l,k,j,i;
double scale,hh,h,g,f;
for (i=n;i>=2;i--) {
l=i-1;
h=scale=0.0;
if (l > 1) {
for (k=1;k<=l;k++)
scale += fabs(a[i][k]);
if (scale == 0.0) /* Skip transformation. */
e[i]=a[i][l];
else {
for (k=1;k<=l;k++) {
a[i][k] /= scale; /* Use scaled a's for transformation. */
h += a[i][k]*a[i][k]; /* Form sigma in h. */
}
f=a[i][l];
g=(f >= 0.0 ? -sqrt(h) : sqrt(h));
e[i]=scale*g;
h -= f*g; /* Now h is equation (11.2.4). */
a[i][l]=f-g; /* Store u in the ith row of a. */
f=0.0;
for (j=1;j<=l;j++) {
/* Next statement can be omitted if eigenvectors not wanted */
a[j][i]=a[i][j]/h; /* Store u/H in ith column of a. */
g=0.0; /* Form an element of A.u in g. */
for (k=1;k<=j;k++)
g += a[j][k]*a[i][k];
for (k=j+1;k<=l;k++)
g += a[k][j]*a[i][k];
e[j]=g/h; /* Form element of p in temporarily unused element of e. */
f += e[j]*a[i][j];
}
hh=f/(h+h); /* Form K, equation (11.2.11). */
for (j=1;j<=l;j++) { /* Form q and store in e overwriting p. */
f=a[i][j];
e[j]=g=e[j]-hh*f;
for (k=1;k<=j;k++) /* Reduce a, equation (11.2.13). */
a[j][k] -= (f*e[k]+g*a[i][k]);
}
}
} else
e[i]=a[i][l];
d[i]=h;
}
/* Next statement can be omitted if eigenvectors not wanted */
d[1]=0.0;
e[1]=0.0;
/* Contents of this loop can be omitted if eigenvectors not
wanted except for statement d[i]=a[i][i]; */
for (i=1;i<=n;i++) { /* Begin accumulation of transformation matrices. */
l=i-1;
if (d[i]) { /* This block skipped when i=1. */
for (j=1;j<=l;j++) {
g=0.0;
for (k=1;k<=l;k++) /* Use u and u/H stored in a to form P.Q. */
g += a[i][k]*a[k][j];
for (k=1;k<=l;k++)
a[k][j] -= g*a[k][i];
}
}
d[i]=a[i][i]; /* This statement remains. */
a[i][i]=1.0; /* Reset row and column of a to identity matrix for next iteration. */
for (j=1;j<=l;j++) a[j][i]=a[i][j]=0.0;
}
}
/******************************************************************************/
void tqli(double d[], double e[], int n, double **z)
/*******************************************************************************
QL algorithm with implicit shifts, to determine the eigenvalues and eigenvectors
of a real, symmetric, tridiagonal matrix, or of a real, symmetric matrix
previously reduced by tred2 sec. 11.2. On input, d[1..n] contains the diagonal
elements of the tridiagonal matrix. On output, it returns the eigenvalues. The
vector e[1..n] inputs the subdiagonal elements of the tridiagonal matrix, with
e[1] arbitrary. On output e is destroyed. When finding only the eigenvalues,
several lines may be omitted, as noted in the comments. If the eigenvectors of
a tridiagonal matrix are desired, the matrix z[1..n][1..n] is input as the
identity matrix. If the eigenvectors of a matrix that has been reduced by tred2
are required, then z is input as the matrix output by tred2. In either case,
the kth column of z returns the normalized eigenvector corresponding to d[k].
*******************************************************************************/
{
double pythag(double a, double b);
int m,l,iter,i,k;
double s,r,p,g,f,dd,c,b;
for (i=2;i<=n;i++) e[i-1]=e[i]; /* Convenient to renumber the elements of e. */
e[n]=0.0;
for (l=1;l<=n;l++) {
iter=0;
do {
for (m=l;m<=n-1;m++) { /* Look for a single small subdiagonal element to split the matrix. */
dd=fabs(d[m])+fabs(d[m+1]);
if ((double)(fabs(e[m])+dd) == dd) break;
}
if (m != l) {
if (iter++ == 30) printf("Too many iterations in tqli");
g=(d[l+1]-d[l])/(2.0*e[l]); /* Form shift. */
r=pythag(g,1.0);
g=d[m]-d[l]+e[l]/(g+SIGN(r,g)); /* This is dm - ks. */
s=c=1.0;
p=0.0;
for (i=m-1;i>=l;i--) { /* A plane rotation as in the original QL, followed by Givens */
f=s*e[i]; /* rotations to restore tridiagonal form. */
b=c*e[i];
e[i+1]=(r=pythag(f,g));
if (r == 0.0) { /* Recover from underflow. */
d[i+1] -= p;
e[m]=0.0;
break;
}
s=f/r;
c=g/r;
g=d[i+1]-p;
r=(d[i]-g)*s+2.0*c*b;
d[i+1]=g+(p=s*r);
g=c*r-b;
/* Next loop can be omitted if eigenvectors not wanted */
for (k=1;k<=n;k++) { /* Form eigenvectors. */
f=z[k][i+1];
z[k][i+1]=s*z[k][i]+c*f;
z[k][i]=c*z[k][i]-s*f;
}
}
if (r == 0.0 && i >= l) continue;
d[l] -= p;
e[l]=g;
e[m]=0.0;
}
} while (m != l);
}
}
/******************************************************************************/
double pythag(double a, double b)
/*******************************************************************************
Computes (a2 + b2)1/2 without destructive underflow or overflow.
*******************************************************************************/
{
double absa,absb;
absa=fabs(a);
absb=fabs(b);
if (absa > absb) return absa*sqrt(1.0+(absb/absa)*(absb/absa));
else return (absb == 0.0 ? 0.0 : absb*sqrt(1.0+(absa/absb)*(absa/absb)));
}
#define NP 3
#define TINY 1.0e-6
double sqrt(double x)
{
union
{
int i;
double x;
} u;
u.x = x;
u.i = (1<<29) + (u.i >> 1) - (1<<22);
return u.x;
}
int main()
{
int i,j,k;
double ze[NP][NP] = { {2, 0, 0}, {0, 4, 0}, {0, 0, 2} } ;
double **a;
double *d,*e,*f;
d=dvector(1,NP);
e=dvector(1,NP);
f=dvector(1,NP);
a=dmatrix(1,NP,1,NP);
for (i=1;i<=NP;i++)
for (j=1;j<=NP;j++) a[i][j]=ze[i-1][j-1];
tred2(a,NP,d,e);
tqli(d,e,NP,a);
printf("\nEigenvectors for a real symmetric matrix:\n");
for (i=1;i<=NP;i++) {
for (j=1;j<=NP;j++) {
f[j]=0.0;
for (k=1;k<=NP;k++)
f[j] += (ze[j-1][k-1]*a[k][i]);
}
printf("%s %3d %s %10.6f\n","\neigenvalue",i," =",d[i]);
printf("%11s %14s %9s\n","vector","mtrx*vect.","ratio");
for (j=1;j<=NP;j++) {
if (fabs(a[j][i]) < TINY)
printf("%12.6f %12.6f %12s\n",
a[j][i],f[j],"div. by 0");
else
printf("%12.6f %12.6f %12.6f\n",
a[j][i],f[j],f[j]/a[j][i]);
}
}
//free_dmatrix(a,1,NP,1,NP);
//free_dvector(f,1,NP);
//free_dvector(e,1,NP);
//free_dvector(d,1,NP);
return 0;
}
Output:
Eigenvectors for a real symmetric matrix:
eigenvalue 1 = 2.000000
vector mtrx*vect. ratio
1.000000 2.000000 2.000000
0.000000 0.000000 div. by 0
0.000000 0.000000 div. by 0
eigenvalue 2 = 4.000000
vector mtrx*vect. ratio
0.000000 0.000000 div. by 0
1.000000 4.000000 4.000000
0.000000 0.000000 div. by 0
eigenvalue 3 = 2.000000
vector mtrx*vect. ratio
0.000000 0.000000 div. by 0
0.000000 0.000000 div. by 0
1.000000 2.000000 2.000000
I hope it finaly helps to clarify confusion regarding the data preparation for tqli.

Solving linear equation with LAPACKE

I'm trying to solve some linear equation (which is symmetrical, tridiagonal and positive). I have to use LAPACKE. My code is as follows:
#include <lapacke.h>
#include <stdio.h>
void print_mtrx(double * mtrx, int n, int m)
{
int i, j;
for(i = 0; i < n; i++)
{
for(j = 0; j < m; j++)
{
printf("%f ", mtrx[i*m+j]);
}
printf("\n");
}
printf("\n");
}
int main()
{
double matrix[5*5] = {
2, 0, 0, 0, 0,
0, 2, 0, 0, 0,
0, 0, 2, 0, 0,
0, 0, 0, 2, 0,
0, 0, 0, 0, 2
};
double rozw[5] = {1,2,3,4,5};
double matrix2[5*5] = {
7, 0, 0, 0, 0,
0, 7, 0, 0, 0,
0, 0, 7, 0, 0,
0, 0, 0, 7, 0,
0, 0, 0, 0, 7
};
LAPACKE_dptsv(LAPACK_COL_MAJOR, 5, 5, matrix, matrix2, rozw, 5);
print_mtrx(matrix, 5, 5);
print_mtrx(matrix2, 5, 5);
print_mtrx(rozw, 5, 1);
}
LAPACKE's function seems to do nothing, without any errors. The main problem is, I have no idea what do the function parameters stand for. I've searched long, but there is no real documentation. Here is what I managed to found or guess:
int matrix_order -- LAPACK_COL_MAJOR or LAPACK_ROW_MAJOR, how is matrix represented in memory
lapack_int n -- size of matrix (i.e. number of columns)
lapack_int nrhs -- not sure, maybe size of vector b
double* d -- matrix of equations
double* e -- no idea.
double* b -- vector of solutions of equations from d
lapack_int ldb -- leading direction of b (wherefore? is it not identical to nrhs, which itself is identical to n?)
How can I find real meaning of these arguments? How can I make my code work?
So one must to look into documentation of pure LAPACK (http://www.netlib.org/lapack/explore-html/d0/dea/dptsv_8f.html#af1bd4c731915bd8755a4da8086fd79a8), and also ignore incorrect (in case of LAPACKE) remark that LDB be greater or equal max(1,N).
The correct program is as follows:
#include <lapacke.h>
#include <stdio.h>
void print_mtrx(double * mtrx, int n, int m)
{
int i, j;
for(i = 0; i < n; i++)
{
for(j = 0; j < m; j++)
{
printf("%f ", mtrx[i*m+j]);
}
printf("\n");
}
printf("\n");
}
int main()
{
double diagonal[5] = {5,1,5,1,5};
double subdiagonal[4] = {0,0,0,0};
double solution[5] = {1,2,3,4,5};
LAPACKE_dptsv(LAPACK_ROW_MAJOR, 5 /*size of matrix*/, 1 /*number of columns in solution*/,
diagonal, subdiagonal, solution, 1 /*leading dimension of solution vector*/);
print_mtrx(solution, 5, 1);
}
When it comes to documentation for BLAS and/or LAPACK, Intel is probably the most comprehensive out there. You can look up the docs for ?ptsv, which explains what each parameter is for.
(Hint: when searching for a BLAS or LAPACK in Google, be sure to drop the s/d/c/z prefix.)
Here's the relevant snippet:
The routine solves for X the real or complex system of linear equations A*X = B, where A is an n-by-n symmetric/Hermitian positive-definite tridiagonal matrix, the columns of matrix B are individual right-hand sides, and the columns of X are the corresponding solutions.
A is factored as A = L*D*LT (real flavors) or A = L*D*LH (complex flavors), and the factored form of A is then used to solve the system of equations A*X = B.
Input Parameters
n: The order of matrix A; n ≥ 0.
nrhs: The number of right-hand sides, the number of columns in B; nrhs ≥ 0.
d: Array, dimension at least max(1, n). Contains the diagonal elements of the tridiagonal matrix A.
e, b: Arrays: e(n - 1), b(ldb,*). The array e contains the (n - 1) subdiagonal elements of A. The array b contains the matrix B whose columns are the right-hand sides for the systems of equations. The second dimension of b must be at least max(1,nrhs).
ldb: The leading dimension of b; ldb ≥ max(1, n).
Output Parameters
d: Overwritten by the n diagonal elements of the diagonal matrix D from the L*D*LT (real) / L*D*LH (complex) factorization of A.
e: Overwritten by the (n - 1) subdiagonal elements of the unit bidiagonal factor L from the factorization of A.
b: Overwritten by the solution matrix X.
info: If info = 0, the execution is successful. If info = -i, the i-th parameter had an illegal value. If info = i, the leading minor of order i (and therefore the matrix A itself) is not positive-definite, and the solution has not been computed. The factorization has not been completed unless i = n.

Line fit from an array of 2d vectors

I have a problem in some C code, I assume it belonged here over the Mathematics exchange.
I have an array of changes in x and y position generated by a user dragging a mouse, how could I determine if a straight line was drawn or not.
I am currently using linear regression, is there a better(more efficient) way to do this?
EDIT:
Hough transformation attempt:
#define abSIZE 100
#define ARRAYSIZE 10
int A[abSIZE][abSIZE]; //points in the a-b plane
int dX[10] = {0, 10, 13, 8, 20, 18, 19, 22, 12, 23};
int dY[10] = {0, 2, 3, 1, -1, -2, 0, 0, 3, 1};
int absX[10]; //absolute positions
int absY[10];
int error = 0;
int sumx = 0, sumy = 0, i;
//Convert deltas to absolute positions
for (i = 0; i<10; i++) {
absX[i] = sumx+=dX[i];
absY[i] = sumy+=dY[i];
}
//initialise array to zero
int a, b, x, y;
for(a = -abSIZE/2; a < abSIZE/2; a++) {
for(b = -abSIZE/2; b< abSIZE/2; b++) {
A[a+abSIZE/2][b+abSIZE/2] = 0;
}
}
//Hough transform
int aMax = 0;
int bMax = 0;
int highest = 0;
for(i=0; i<10; i++) {
x = absX[i];
y = absX[i];
for(a = -abSIZE/2; a < abSIZE/2; a++) {
for(b = -abSIZE/2; b< abSIZE/2; b++) {
if (a*x + b == y) {
A[a+abSIZE/2][b+abSIZE/2] += 1;
if (A[a+abSIZE/2][b+abSIZE/2] > highest) {
highest++; //highest = A[a+abSIZE/2][b+abSIZE/2]
aMax = a;
bMax = b;
}
}
}
}
}
printf("Line is Y = %d*X + %d\n",aMax,bMax);
//Calculate MSE
int e;
for (i = 0; i < ARRAYSIZE; i++) {
e = absY[i] - (aMax * absX[i] + bMax);
e = (int) pow((double)e, 2);
error += e;
}
printf("error is: %d\n", error);
Though linear regression sounds like a perfectly reasonable way to solve the task, here's another suggestion: Hough transform, which might be somewhat more robust against outliers. Here is a very rough sketch of how this can be applied:
initialize a large matrix A with zeros
transform your deltas to some absolute coordinates (x, y) in a x-y-plane (e.g. start with (0,0))
for each point
there are non-unique parameters a and b such that a*x + b = y. All such points (a,b) define a straight line in the a-b-plane
draw this "line" in the a-b-plane by adding ones to the corresponding cells in A, which represents the quantized plane
now you can find a maximum in the a-b-plane-matrix A, which will correspond to the parameters (a, b) of the straight line in the x-y-plane that has most support by the original points
finally, calculate MSE to the original points and decide with some threshold if the move was a straight line
More details e.g. here:
http://homepages.inf.ed.ac.uk/rbf/CVonline/LOCAL_COPIES/MARSHALL/node32.html
Edit: here's a quote from Wikipedia that explains why it's better to use a different parametrization to deal with vertical lines (where a would become infinite in ax+b=y):
However, vertical lines pose a problem. They are more naturally described as x = a and would give rise to unbounded values of the slope parameter m. Thus, for computational reasons, Duda and Hart proposed the use of a different pair of parameters, denoted r and theta, for the lines in the Hough transform. These two values, taken in conjunction, define a polar coordinate.
Thanks to Zaw Lin for pointing this out.

How do I calculate the degree between the x=0 line and the line from the (0, 0) to the point (j, i)?

I have to return the angle in degrees and it should be in the range between 0 and 360 (inclusively). I have to use the functions in the math.h library, acos(), asin(), and atan(). In my instructions it says that depending on the quadrant that the point is in, the computation may be different.
Note: I will be using this function to draw an image later on, so I was thinking that the reason they say point (j, i) is because they are referring to the number of rows and columns, respectively. But I'm a little confused on how they are plotting the points. Is it the way you would plot on a coordinate plane in math? (j being the x coordinate and i being a y coordinate)
I've attempted a solution, but I got a little confused by the idea that depending on the quadrants, the computations would be different. I was thinking that I could use acos() for all four quadrants if I found the distance from the origin to the point (j, i) and if I just used the adjacent side which would just be the j coordinate. (If j is equivalent to the x coordinate when plotting the points)
float getAngle(int i, int j) {
float distance, result, radians, degrees;
distance = sqrt((j * j) + (i * i));
if (j > 0 && i > 0) {
result = cos(distance/j) * (1.0);
radians = acos(result);
degrees = radians * (180/3.14);
}
}
Thank you in advance to everyone who posts advice up! I really appreciate it!
#include <stdio.h>
#include <math.h>
#define PI 3.14159265358979323
#if 0
double getAngle(int x, int y) {
//angle(degree) from (0, 0) to (x, y) by acos
double distance, cos_v, radians, degrees;
distance = sqrt((x * x) + (y * y));
cos_v = x / distance;
radians = acos(cos_v);
if(y<0)
radians=2.0*PI-radians;
degrees = radians * (180.0/PI);
}
#else
double getAngle(int x, int y) {
//angle(degree) from (0, 0) to (x, y) by atan2
//(x ==0 && y==0) is error
double radians, degrees;
radians = atan2(y, x);
if(radians<0)
radians += 2.0*PI;
degrees = radians * (180.0/PI);
}
#endif
int main(void){
int data[][2] = {
//(x, y)
{ 1, 0}, { 1, 1}, {0, 1}, {-1, 1},
{-1, 0}, {-1, -1}, {0, -1}, { 1, -1}
};
int i;
for(i=0;i<8;++i)
printf("%g\n", getAngle(data[i][0], data[i][1]));
return 0;
}

Resources