Related
I am writing a simple newton method
x_(n+1) = x_n - f(x_n) / f_prime(x_n)
to find the roots (can be a real number or a complex number) of a quadratic function:
f(x) = a*x*x + b*x + c
(a, b, c are given constants and are all real numbers). I know Newton method will fail if the start point or some iteration point in the loop has a zero derivative. I want to use a if statement inside my for/while loop to avoid this situation. Does Julia have something like stop 0 syntax in Fortran ?
The generic Newton's Method root-finding code:
function newton_root_finding(f, f_diff, x0, rtol=1e-8, atol=1e-8)
f_x0 = f(x0)
f_diff_x0 = f_diff(x0)
x1 = x0 - f_x0 / f_diff_x0
f_diff_x1 = f_diff(x1)
#assert abs(f_diff_x0) > atol + rtol * abs(f_diff_x0) "Zero derivative. No solution found."
while abs(f_x0) > atol + rtol * (abs(f_x0))
x0 = x1
f_x0 = f(x0)
f_diff_x0 = f_diff(x0)
x1 = x0 - f_x0 / f_diff_x0
end
return x1
end
function quadratic_func(x)
a = 1.0
b = 0.0
c = 2.0
return a*x*x + b*x + c
end
function quadratic_func_diff(x)
a = 1.0
b = 0.0
c = 2.0
return 2.0*a*x + 1.0*b + 0.0*c
end
newton_root_finding(quadratic_func, quadratic_func_diff, 1.0 + 0.5im)
In the above code I used a #assert macro to make it happens, but I don't want to use any macro. I want to use a if statement inside my while loop to halt it. Another thing I've noticed is that if I change to #assert abs(f_diff_x0) != 0 this test will be ignored. Is that because of some round-off errors that "zero derivative" doesn't exactly equal to 0?
The way to exit from the inside of a loop in general is a break statement; a return fulfills the same purpose, because it just exits the whole function.
For the comparisons you can use Base.isapprox(x, y; atol=atol, rtol=rtol). It's documentation starts with:
Inexact equality comparison: true if norm(x-y) <= max(atol, rtol*max(norm(x), norm(y))).
norm falls back to abs for numbers. And I think you might have a bug in both comparisons, always comparing the value at x0 to itself.
As for the breaking on zero derivatives, an #assert is, I think, appropriate here: if you get zero derivative, you don't stop iteration and return the result, but you throw an error to signify an infeasible condition. I'd thus write your function as follows:
function newton_root_finding(f, ∂f, x0, rtol=1e-8, atol=1e-8)
x_old = x0
y_old = f(x0)
while true
df_old = ∂f(x_old)
#assert !isapprox(df_old, 0, rtol=rtol, atol=atol) "Zero derivative. No solution found."
x_new = x_old - y_old / df_old
y_new = f(x_new)
isapprox(y_old, y_new, rtol=rtol, atol=atol) && return x_new
x_old, y_old = x_new, y_new
end
end
This returns 3.357392012620626e-26 + 1.4142135623730951im on your test case, approximately sqrt(2)im.
To address your first question, you can use break to exit the while loop, like
function test()
i = 0
while true
i += 1
if i > 10
break
end
end
return i
end
As to your second question, when comparing floating point numbers it is often better to use isapprox (provide an atol if you compare against zero) instead of == or !=.
Hi have made this code to plot a function.
I need to mark with an red X all the crossings between x = 0 and the blue wave line in the graph.
I have made some tries but with '-xr' in the plot function but it places X marks out of the crossings.
Anyone knows how to do it. Many thanks.
Code:
% entrada
a = input('Introduza o valor de a: ');
% ficheiro fonte para a função
raizes;
% chamada à função
x = 0:.1:50;
or = x;
or(:) = 0;
h = #(x) cos(x);
g = #(x) exp(a*x)-1;
f = #(x) h(x) - g(x);
zeros = fzero(f,0);
plot(x,f(x));
hold on
plot(zeros,f(zeros),'-xr')
hold off
Graph (it only marks one zero, i need all the zero crossings):
As mentioned in the comments above, you need to look for the zeros of your function before you can plot them. You can do this mathematically (in this case set f(x) = g(x) and solve for x) or you can do this analytically with something like fsolve.
If you read the documentation for fsolve, you will see that it searches for the zero closest to the provided x0 if passed a scalar or the first zero if passed an interval. What we can do for a quick attempt at a solution is to pass our x values into fsolve as initial guesses and filter out the unique values.
% Set up sample data
a = .05;
x = 0:.1:50;
% Set up equations
h = #(x) cos(x);
g = #(x) exp(a*x)-1;
f = #(x) h(x) - g(x);
% Find zeros of f(x)
crossingpoints = zeros(length(x), 1); % Initialize array
for ii = 1:length(x) % Use x data points as guesses for fzero
try
crossingpoints(ii) = fzero(f, x(ii)); % Find zero closest to guess
end
end
crossingpoints(crossingpoints < 0) = []; % Throw out zeros where x < 0
% Find unique zeros
tol = 10^-8;
crossingpoints = sort(crossingpoints(:)); % Sort data for easier diff
temp = false(size(crossingpoints)); % Initialize testing array
% Find where the difference between 'zeros' is less than or equal to the
% tolerance and throw them out
temp(1:end-1) = abs(diff(crossingpoints)) <= tol;
crossingpoints(temp) = [];
% Sometimes catches beginning of the data set, filter it out if this happens
if abs(f(crossingpoints(1))) >= (0 + tol)
crossingpoints(1) = [];
end
% Plot data
plot(x, f(x))
hold on
plot(crossingpoints, f(crossingpoints), 'rx')
hold off
grid on
axis([0 20 -2 2]);
Which gives us the following:
Note that due to errors arising from floating point arithmetic we have to utilize a tolerance to filter our zeros rather than utilizing a function like unique.
My first Fortran lesson is to plot the probability density function of the radial Sturmian functions. In case you are interested, the radial Sturmian functions are used to graph the momentum space eigenfunctions for the hydrogen atom.
In order to produce these radial functions, one needs to first produce some polynomials called the Gegenbauer polynomials, denoted
Cba(x),
where a and b should be stacked atop each other. One needs these polynomials because the Sturmians (let's call them R_n,l) are defined like so,
R_n,l(p) = N pl⁄(p2 + k2)l+2 Cn - l - 1l + 1(p2 - k2⁄p2 + k2),
where N is a normalisation constant, p is the momentum, n is the principle quantum number, l is the angular momentum and k is a constant. The normalisation constant is there so that when I come to square this function, it will produce a probability distribution for the momentum of the electron in a hydrogen atom.
Gegenbauer polynomials are generated using the following recurrence relation:
Cnl(x) = 1⁄n[2(l + n - 1) x Cn - 1l(x) - (2l + n - 2)Cn - 2l(x)],
with C0l(x) = 1 and C1l(x) = 2lx, as you may have noticed, l is fixed but n is not. At the start of my program, I will specify both l and n and work out the Gegenbauer polynomial I need for the radial function I wish to plot.
The problems I am having with my code at the moment are all in my subroutine for working out the value of the Gegenbauer polynomial Cn-l-1l+1(p2 - k2⁄p2 + k2) for incremental values of p between 0 and 3. I keep getting the error
Unclassified statement at (1)
but I cannot see what the issue is.
program Radial_Plot
implicit none
real, parameter :: pi = 4*atan(1.0)
integer, parameter :: top = 1000, l = 50, n = 100
real, dimension(1:top) :: x, y
real increment
real :: a=0.0, b = 2.5, k = 0.3
integer :: i
real, dimension(1:top) :: C
increment = (b-a)/(real(top)-1)
x(1) = 0.0
do i = 2, top
x(i) = x(i-1) + increment
end do
Call Gegenbauer(top, n, l, k, C)
y = x*C
! y is the function that I shall be plotting between values a and b.
end program Radial_Plot
Subroutine Gegenbauer(top1, n1, l1, k1, CSub)
! This subroutine is my attempt to calculate the Gegenbauer polynomials evaluated at a certain number of values between c and d.
implicit none
integer :: top1, i, j, n1, l1
real :: k1, increment1, c, d
real, dimension(1:top1) :: x1
real, dimension(1:n1 - l1, 1:top1) :: C1
real, dimension(1:n1 - l1) :: CSub
c = 0.0
d = 3.0
k1 = 0.3
n1 = 50
l1 = 25
top1 = 1000
increment1 = (d - c)/(real(top1) - 1)
x1(1) = 0.0
do i = 2, top1
x1(i) = x1(i-1) + increment1
end do
do j = 1, top1
C1(1,j) = 1
C1(2,j) = 2(l1 + 1)(x1(i)^2 - k1^2)/(x1(i)^2 + k1^2)
! All the errors occurring here are all due to, and I quote, 'Unclassifiable statement at (1)', I can't see what the heck I have done wrong.
do i = 3, n1 - l1
C1(i,j) = 2(((l1 + 1)/n1) + 1)(x1(i)^2 - k1^2)/(x1(i)^2 + k1^2)C1(i,j-1) - ((2(l1+1)/n1) + 1)C1(i,j-2)
end do
CSub(j) = Cn(n1 - l1,j)^2
end do
return
end Subroutine Gegenbauer
As francesalus correctly pointed out, the problem is because you use ^ instead of ** for exponentiation. Additionally, you do not put * between the terms you are multiplying.
C1(1,j) = 1
C1(2,j) = 2*(l1 + 1)*(x1(i)**2 - k1**2)/(x1(i)**2 + k1**2)
do i = 3, n1 - l1
C1(i,j) = 2 * (((l1 + 1)/n1) + 1) * (x1(i)**2 - k1**2) / &
(x1(i)**2 + k1**2)*C1(i,j-1) - ((2(l1+1)/n1) + 1) * &
C1(i,j-2)
end do
CSub(j) = Cn(n1 - l1,j)**2
Since you are beginning I have some advice. Learn to put all subroutines and functions to modules (unless they are internal). There is no reason for the return statement at the and of the subroutine, similarly as a stop statement isn't necessary at the and of the program.
How can I find the line of intersection between two planes?
I know the mathematics idea, and I did the cross product between the the planes normal vectors
but how to get the line from the resulted vector programmatically
The equation of the plane is ax + by + cz + d = 0, where (a,b,c) is the plane's normal, and d is the distance to the origin. This means that every point (x,y,z) that satisfies that equation is a member of the plane.
Given two planes:
P1: a1x + b1y + c1z + d1 = 0
P2: a2x + b2y + c2z + d2 = 0
The intersection between the two is the set of points that verifies both equations. To find points along this line, you can simply pick a value for x, any value, and then solve the equations for y and z.
y = (-c1z -a1x -d1) / b1
z = ((b2/b1)*(a1x+d1) -a2x -d2)/(c2 - c1*b2/b1)
If you make x=0, this gets simpler:
y = (-c1z -d1) / b1
z = ((b2/b1)*d1 -d2)/(c2 - c1*b2/b1)
Finding the line between two planes can be calculated using a simplified version of the 3-plane intersection algorithm.
The 2'nd, "more robust method" from bobobobo's answer references the 3-plane intersection.
While this works well for 2 planes (where the 3rd plane can be calculated using the cross product of the first two), the problem can be further reduced for the 2-plane version.
No need to use a 3x3 matrix determinant,instead we can use the squared length of the cross product between the first and second plane (which is the direction of the 3'rd plane).
No need to include the 3rd planes distance,(calculating the final location).
No need to negate the distances.Save some cpu-cycles by swapping the cross product order instead.
Including this code-example, since it may not be immediately obvious.
// Intersection of 2-planes: a variation based on the 3-plane version.
// see: Graphics Gems 1 pg 305
//
// Note that the 'normal' components of the planes need not be unit length
bool isect_plane_plane_to_normal_ray(
const Plane& p1, const Plane& p2,
// output args
Vector3f& r_point, Vector3f& r_normal)
{
// logically the 3rd plane, but we only use the normal component.
const Vector3f p3_normal = p1.normal.cross(p2.normal);
const float det = p3_normal.length_squared();
// If the determinant is 0, that means parallel planes, no intersection.
// note: you may want to check against an epsilon value here.
if (det != 0.0) {
// calculate the final (point, normal)
r_point = ((p3_normal.cross(p2.normal) * p1.d) +
(p1.normal.cross(p3_normal) * p2.d)) / det;
r_normal = p3_normal;
return true;
}
else {
return false;
}
}
Adding this answer for completeness, since at time of writing, none of the answers here contain a working code-example which directly addresses the question.
Though other answers here already covered the principles.
Finding a point on the line
To get the intersection of 2 planes, you need a point on the line and the direction of that line.
Finding the direction of that line is really easy, just cross the 2 normals of the 2 planes that are intersecting.
lineDir = n1 × n2
But that line passes through the origin, and the line that runs along your plane intersections might not. So, Martinho's answer provides a great start to finding a point on the line of intersection (basically any point that is on both planes).
In case you wanted to see the derivation for how to solve this, here's the math behind it:
First let x=0. Now we have 2 unknowns in 2 equations instead of 3 unknowns in 2 equations (we arbitrarily chose one of the unknowns).
Then the plane equations are (A terms were eliminated since we chose x=0):
B1y + C1z + D1 = 0
B2y + C2z + D2 = 0
We want y and z such that those equations are both solved correctly (=0) for the B1, C1 given.
So, just multiply the top eq by (-B2/B1) to get
-B2y + (-B2/B1)*C1z + (-B2/B1)*D1 = 0
B2y + C2z + D2 = 0
Add the eqs to get
z = ( (-B2/B1)*D1 - D2 ) / (C2 * B2/B1)*C1)
Throw the z you find into the 1st equation now to find y as
y = (-D1 - C1z) / B1
Note the best variable to make 0 is the one with the lowest coefficients, because it carries no information anyway. So if C1 and C2 were both 0, choosing z=0 (instead of x=0) would be a better choice.
The above solution can still screw up if B1=0 (which isn't that unlikely). You could add in some if statements that check if B1=0, and if it is, be sure to solve for one of the other variables instead.
Solution using intersection of 3 planes
From user's answer, a closed form solution for the intersection of 3 planes was actually in Graphics Gems 1. The formula is:
P_intersection = (( point_on1 • n1 )( n2 × n3 ) + ( point_on2 • n2 )( n3 × n1 ) + ( point_on3 • n3 )( n1 × n2 )) / det(n1,n2,n3)
Actually point_on1 • n1 = -d1 (assuming you write your planes Ax + By + Cz + D=0, and not =-D). So, you could rewrite it as:
P_intersection = (( -d1 )( n2 × n3 ) + ( -d2 )( n3 × n1 ) + ( -d3 )( n1 × n2 )) / det(n1,n2,n3)
A function that intersects 3 planes:
// Intersection of 3 planes, Graphics Gems 1 pg 305
static Vector3f getIntersection( const Plane& plane1, const Plane& plane2, const Plane& plane3 )
{
float det = Matrix3f::det( plane1.normal, plane2.normal, plane3.normal ) ;
// If the determinant is 0, that means parallel planes, no intn.
if( det == 0.f ) return 0 ; //could return inf or whatever
return ( plane2.normal.cross( plane3.normal )*-plane1.d +
plane3.normal.cross( plane1.normal )*-plane2.d +
plane1.normal.cross( plane2.normal )*-plane3.d ) / det ;
}
Proof it works (yellow dot is intersection of rgb planes here)
Getting the line
Once you have a point of intersection common to the 2 planes, the line just goes
P + t*d
Where P is the point of intersection, t can go from (-inf, inf), and d is the direction vector that is the cross product of the normals of the two original planes.
The line of intersection between the red and blue planes looks like this
Efficiency and stability
The "robust" (2nd way) takes 48 elementary ops by my count, vs the 36 elementary ops that the 1st way (isolation of x,y) uses. There is a trade off between stability and # computations between these 2 ways.
It'd be pretty catastrophic to get (0,inf,inf) back from a call to the 1st way in the case that B1 was 0 and you didn't check. So adding in if statements and making sure not to divide by 0 to the 1st way may give you the stability at the cost of code bloat, and the added branching (which might be quite expensive). The 3 plane intersection method is almost branchless and won't give you infinities.
This method avoids division by zero as long as the two planes are not parallel.
If these are the planes:
A1*x + B1*y + C1*z + D1 = 0
A2*x + B2*y + C2*z + D2 = 0
1) Find a vector parallel to the line of intersection. This is also the normal of a 3rd plane which is perpendicular to the other two planes:
(A3,B3,C3) = (A1,B1,C1) cross (A2,B2,C2)
2) Form a system of 3 equations. These describe 3 planes which intersect at a point:
A1*x1 + B1*y1 + C1*z1 + D1 = 0
A2*x1 + B2*y1 + C2*z1 + D2 = 0
A3*x1 + B3*y1 + C3*z1 = 0
3) Solve them to find x1,y1,z1. This is a point on the line of intersection.
4) The parametric equations of the line of intersection are:
x = x1 + A3 * t
y = y1 + B3 * t
z = z1 + C3 * t
The determinant-based approach is neat, but it's hard to follow why it works.
Here's another way that's more intuitive.
The idea is to first go from the origin to the closest point on the first plane (p1), and then from there go to the closest point on the line of intersection of the two planes. (Along a vector that I'm calling v below.)
Given
=====
First plane: n1 • r = k1
Second plane: n2 • r = k2
Working
=======
dir = n1 × n2
p1 = (k1 / (n1 • n1)) * n1
v = n1 × dir
pt = LineIntersectPlane(line = (p1, v), plane = (n2, k2))
LineIntersectPlane
==================
#We have n2 • (p1 + lambda * v) = k2
lambda = (k2 - n2 • p1) / (n2 • v)
Return p1 + lambda * v
Output
======
Line where two planes intersect: (pt, dir)
This should give the same point as the determinant-based approach. There's almost certainly a link between the two. At least the denominator, n2 • v, is the same, if we apply the "scalar triple product" rule. So these methods are probably similar as far as condition numbers go.
Don't forget to check for (almost) parallel planes. For example: if (dir • dir < 1e-8) should work well if unit normals are used.
You can find the formula for the intersection line of two planes in this link.
P1: a1x + b1y + c1z = d1
P2: a2x + b2y + c2z = d2
n1=(a1,b1,c1); n2=(a2,b2,c2); n12=Norm[Cross[n1,n2]]^2
If n12 != 0
a1 = (d1*Norm[n2]^2 - d2*n1.n2)/n12;
a2 = (d2*Norm[n1]^2 - d1*n1.n2)/n12;
P = a1 n1 + a2 n2;
(*formula for the intersection line*)
Li[t_] := P + t*Cross[n1, n2];
The cross product of the line is the direction of the intersection line. Now you need a point in the intersection.
You can do this by taking a point on the cross product, then subtracting Normal of plane A * distance to plane A and Normal of plane B * distance to plane b. Cleaner:
p = Point on cross product
intersection point = ([p] - ([Normal of plane A] * [distance from p to plane A]) - ([Normal of plane B] * [distance from p to plane B]))
Edit:
You have two planes with two normals:
N1 and N2
The cross product is the direction of the Intersection Line:
C = N1 x N2
The class above has a function to calculate the distance between a point and a plane. Use it to get the distance of some point p on C to both planes:
p = C //p = 1 times C to get a point on C
d1 = plane1.getDistance(p)
d2 = plane2.getDistance(p)
Intersection line:
resultPoint1 = (p - (d1 * N1) - (d2 * N2))
resultPoint2 = resultPoint1 + C
I need to programmatically solve a system of linear equations in C, Objective C, or (if needed) C++.
Here's an example of the equations:
-44.3940 = a * 50.0 + b * 37.0 + tx
-45.3049 = a * 43.0 + b * 39.0 + tx
-44.9594 = a * 52.0 + b * 41.0 + tx
From this, I'd like to get the best approximation for a, b, and tx.
Cramer's Rule
and
Gaussian Elimination
are two good, general-purpose algorithms (also see Simultaneous Linear Equations). If you're looking for code, check out GiNaC, Maxima, and SymbolicC++ (depending on your licensing requirements, of course).
EDIT: I know you're working in C land, but I also have to put in a good word for SymPy (a computer algebra system in Python). You can learn a lot from its algorithms (if you can read a bit of python). Also, it's under the new BSD license, while most of the free math packages are GPL.
You can solve this with a program exactly the same way you solve it by hand (with multiplication and subtraction, then feeding results back into the equations). This is pretty standard secondary-school-level mathematics.
-44.3940 = 50a + 37b + c (A)
-45.3049 = 43a + 39b + c (B)
-44.9594 = 52a + 41b + c (C)
(A-B): 0.9109 = 7a - 2b (D)
(B-C): 0.3455 = -9a - 2b (E)
(D-E): 1.2564 = 16a (F)
(F/16): a = 0.078525 (G)
Feed G into D:
0.9109 = 7a - 2b
=> 0.9109 = 0.549675 - 2b (substitute a)
=> 0.361225 = -2b (subtract 0.549675 from both sides)
=> -0.1806125 = b (divide both sides by -2) (H)
Feed H/G into A:
-44.3940 = 50a + 37b + c
=> -44.3940 = 3.92625 - 6.6826625 + c (substitute a/b)
=> -41.6375875 = c (subtract 3.92625 - 6.6826625 from both sides)
So you end up with:
a = 0.0785250
b = -0.1806125
c = -41.6375875
If you plug these values back into A, B and C, you'll find they're correct.
The trick is to use a simple 4x3 matrix which reduces in turn to a 3x2 matrix, then a 2x1 which is "a = n", n being an actual number. Once you have that, you feed it into the next matrix up to get another value, then those two values into the next matrix up until you've solved all variables.
Provided you have N distinct equations, you can always solve for N variables. I say distinct because these two are not:
7a + 2b = 50
14a + 4b = 100
They are the same equation multiplied by two so you cannot get a solution from them - multiplying the first by two then subtracting leaves you with the true but useless statement:
0 = 0 + 0
By way of example, here's some C code that works out the simultaneous equations that you're placed in your question. First some necessary types, variables, a support function for printing out an equation, and the start of main:
#include <stdio.h>
typedef struct { double r, a, b, c; } tEquation;
tEquation equ1[] = {
{ -44.3940, 50, 37, 1 }, // -44.3940 = 50a + 37b + c (A)
{ -45.3049, 43, 39, 1 }, // -45.3049 = 43a + 39b + c (B)
{ -44.9594, 52, 41, 1 }, // -44.9594 = 52a + 41b + c (C)
};
tEquation equ2[2], equ3[1];
static void dumpEqu (char *desc, tEquation *e, char *post) {
printf ("%10s: %12.8lf = %12.8lfa + %12.8lfb + %12.8lfc (%s)\n",
desc, e->r, e->a, e->b, e->c, post);
}
int main (void) {
double a, b, c;
Next, the reduction of the three equations with three unknowns to two equations with two unknowns:
// First step, populate equ2 based on removing c from equ.
dumpEqu (">", &(equ1[0]), "A");
dumpEqu (">", &(equ1[1]), "B");
dumpEqu (">", &(equ1[2]), "C");
puts ("");
// A - B
equ2[0].r = equ1[0].r * equ1[1].c - equ1[1].r * equ1[0].c;
equ2[0].a = equ1[0].a * equ1[1].c - equ1[1].a * equ1[0].c;
equ2[0].b = equ1[0].b * equ1[1].c - equ1[1].b * equ1[0].c;
equ2[0].c = 0;
// B - C
equ2[1].r = equ1[1].r * equ1[2].c - equ1[2].r * equ1[1].c;
equ2[1].a = equ1[1].a * equ1[2].c - equ1[2].a * equ1[1].c;
equ2[1].b = equ1[1].b * equ1[2].c - equ1[2].b * equ1[1].c;
equ2[1].c = 0;
dumpEqu ("A-B", &(equ2[0]), "D");
dumpEqu ("B-C", &(equ2[1]), "E");
puts ("");
Next, the reduction of the two equations with two unknowns to one equation with one unknown:
// Next step, populate equ3 based on removing b from equ2.
// D - E
equ3[0].r = equ2[0].r * equ2[1].b - equ2[1].r * equ2[0].b;
equ3[0].a = equ2[0].a * equ2[1].b - equ2[1].a * equ2[0].b;
equ3[0].b = 0;
equ3[0].c = 0;
dumpEqu ("D-E", &(equ3[0]), "F");
puts ("");
Now that we have a formula of the type number1 = unknown * number2, we can simply work out the unknown value with unknown <- number1 / number2. Then, once you've figured that value out, substitute it into one of the equations with two unknowns and work out the second value. Then substitute both those (now-known) unknowns into one of the original equations and you now have the values for all three unknowns:
// Finally, substitute values back into equations.
a = equ3[0].r / equ3[0].a;
printf ("From (F ), a = %12.8lf (G)\n", a);
b = (equ2[0].r - equ2[0].a * a) / equ2[0].b;
printf ("From (D,G ), b = %12.8lf (H)\n", b);
c = (equ1[0].r - equ1[0].a * a - equ1[0].b * b) / equ1[0].c;
printf ("From (A,G,H), c = %12.8lf (I)\n", c);
return 0;
}
The output of that code matches the earlier calculations in this answer:
>: -44.39400000 = 50.00000000a + 37.00000000b + 1.00000000c (A)
>: -45.30490000 = 43.00000000a + 39.00000000b + 1.00000000c (B)
>: -44.95940000 = 52.00000000a + 41.00000000b + 1.00000000c (C)
A-B: 0.91090000 = 7.00000000a + -2.00000000b + 0.00000000c (D)
B-C: -0.34550000 = -9.00000000a + -2.00000000b + 0.00000000c (E)
D-E: -2.51280000 = -32.00000000a + 0.00000000b + 0.00000000c (F)
From (F ), a = 0.07852500 (G)
From (D,G ), b = -0.18061250 (H)
From (A,G,H), c = -41.63758750 (I)
Take a look at the Microsoft Solver Foundation.
With it you could write code like this:
SolverContext context = SolverContext.GetContext();
Model model = context.CreateModel();
Decision a = new Decision(Domain.Real, "a");
Decision b = new Decision(Domain.Real, "b");
Decision c = new Decision(Domain.Real, "c");
model.AddDecisions(a,b,c);
model.AddConstraint("eqA", -44.3940 == 50*a + 37*b + c);
model.AddConstraint("eqB", -45.3049 == 43*a + 39*b + c);
model.AddConstraint("eqC", -44.9594 == 52*a + 41*b + c);
Solution solution = context.Solve();
string results = solution.GetReport().ToString();
Console.WriteLine(results);
Here is the output:
===Solver Foundation Service Report===
Datetime: 04/20/2009 23:29:55
Model Name: Default
Capabilities requested: LP
Solve Time (ms): 1027
Total Time (ms): 1414
Solve Completion Status: Optimal
Solver Selected: Microsoft.SolverFoundation.Solvers.SimplexSolver
Directives:
Microsoft.SolverFoundation.Services.Directive
Algorithm: Primal
Arithmetic: Hybrid
Pricing (exact): Default
Pricing (double): SteepestEdge
Basis: Slack
Pivot Count: 3
===Solution Details===
Goals:
Decisions:
a: 0.0785250000000004
b: -0.180612500000001
c: -41.6375875
For a 3x3 system of linear equations I guess it would be okay to roll out your own algorithms.
However, you might have to worry about accuracy, division by zero or really small numbers and what to do about infinitely many solutions. My suggestion is to go with a standard numerical linear algebra package such as LAPACK.
Are you looking for a software package that'll do the work or actually doing the matrix operations and such and do each step?
The the first, a coworker of mine just used Ocaml GLPK. It is just a wrapper for the GLPK, but it removes a lot of the steps of setting things up. It looks like you're going to have to stick with the GLPK, in C, though. For the latter, thanks to delicious for saving an old article I used to learn LP awhile back, PDF. If you need specific help setting up further, let us know and I'm sure, me or someone will wander back in and help, but, I think it's fairly straight forward from here. Good Luck!
Template Numerical Toolkit from NIST has tools for doing that.
One of the more reliable ways is to use a QR Decomposition.
Here's an example of a wrapper so that I can call "GetInverse(A, InvA)" in my code and it will put the inverse into InvA.
void GetInverse(const Array2D<double>& A, Array2D<double>& invA)
{
QR<double> qr(A);
invA = qr.solve(I);
}
Array2D is defined in the library.
In terms of run-time efficiency, others have answered better than I. If you always will have the same number of equations as variables, I like Cramer's rule as it's easy to implement. Just write a function to calculate determinant of a matrix (or use one that's already written, I'm sure you can find one out there), and divide the determinants of two matrices.
Personally, I'm partial to the algorithms of Numerical Recipes. (I'm fond of the C++ edition.)
This book will teach you why the algorithms work, plus show you some pretty-well debugged implementations of those algorithms.
Of course, you could just blindly use CLAPACK (I've used it with great success), but I would first hand-type a Gaussian Elimination algorithm to at least have a faint idea of the kind of work that has gone into making these algorithms stable.
Later, if you're doing more interesting linear algebra, looking around the source code of Octave will answer a lot of questions.
From the wording of your question, it seems like you have more equations than unknowns and you want to minimize the inconsistencies. This is typically done with linear regression, which minimizes the sum of the squares of the inconsistencies. Depending on the size of the data, you can do this in a spreadsheet or in a statistical package. R is a high-quality, free package that does linear regression, among a lot of other things. There is a lot to linear regression (and a lot of gotcha's), but as it's straightforward to do for simple cases. Here's an R example using your data. Note that the "tx" is the intercept to your model.
> y <- c(-44.394, -45.3049, -44.9594)
> a <- c(50.0, 43.0, 52.0)
> b <- c(37.0, 39.0, 41.0)
> regression = lm(y ~ a + b)
> regression
Call:
lm(formula = y ~ a + b)
Coefficients:
(Intercept) a b
-41.63759 0.07852 -0.18061
function x = LinSolve(A,y)
%
% Recursive Solution of Linear System Ax=y
% matlab equivalent: x = A\y
% x = n x 1
% A = n x n
% y = n x 1
% Uses stack space extensively. Not efficient.
% C allows recursion, so convert it into C.
% ----------------------------------------------
n=length(y);
x=zeros(n,1);
if(n>1)
x(1:n-1,1) = LinSolve( A(1:n-1,1:n-1) - (A(1:n-1,n)*A(n,1:n-1))./A(n,n) , ...
y(1:n-1,1) - A(1:n-1,n).*(y(n,1)/A(n,n)));
x(n,1) = (y(n,1) - A(n,1:n-1)*x(1:n-1,1))./A(n,n);
else
x = y(1,1) / A(1,1);
end
For general cases, you could use python along with numpy for Gaussian elimination. And then plug in values and get the remaining values.