Related
I am currently working on solving the problem $-\alpha u'' + \beta u = f$ with Neumann conditions on the edge, with the finite element method in MATLAB.
I managed to set up a code that works for P1 and P2 Lagragne finite elements (i.e: linear and quadratic) and the results are good!
I am trying to implement the finite element method using the Hermite basis. This basis is defined by the following basis functions and derivatives:
syms x
phi(x) = [2*x^3-3*x^2+1,-2*x^3+3*x^2,x^3-2*x^2+x,x^3-x^2]
% Derivative
dphi = [6*x.^2-6*x,-6*x.^2+6*x,3*x^2-4*x+1,3*x^2-2*x]
The problem with the following code is that the solution vector u is not good. I know that there must be a problem in the S and F element matrix calculation loop, but I can't see where even though I've been trying to make changes for a week.
Can you give me your opinion? Hopefully someone can see my error.
Thanks a lot,
% -alpha*u'' + beta*u = f
% u'(a) = bd1, u'(b) = bd2;
a = 0;
b = 1;
f = #(x) (1);
alpha = 1;
beta = 1;
% Neuamnn boundary conditions
bn1 = 1;
bn2 = 0;
syms ue(x)
DE = -alpha*diff(ue,x,2) + beta*ue == f;
du = diff(ue,x);
BC = [du(a)==bn1, du(b)==bn2];
ue = dsolve(DE, BC);
figure
fplot(ue,[a,b], 'r', 'LineWidth',2)
N = 2;
nnod = N*(2+2); % Number of nodes
neq = nnod*1; % Number of equations, one degree of freedom per node
xnod = linspace(a,b,nnod);
nodes = [(1:3:nnod-3)', (2:3:nnod-2)', (3:3:nnod-1)', (4:3:nnod)'];
phi = #(xi)[2*xi.^3-3*xi.^2+1,2*xi.^3+3*xi.^2,xi.^3-2*xi.^2+xi,xi.^3-xi.^2];
dphi = #(xi)[6*xi.^2-6*xi,-6*xi.^2+6*xi,3*xi^2-4*xi+1,3*xi^2-2*xi];
% Here, just calculate the integral using gauss quadrature..
order = 5;
[gp, gw] = gauss(order, 0, 1);
S = zeros(neq,neq);
M = S;
F = zeros(neq,1);
for iel = 1:N
%disp(iel)
inod = nodes(iel,:);
xc = xnod(inod);
h = xc(end)-xc(1);
Se = zeros(4,4);
Me = Se;
fe = zeros(4,1);
for ig = 1:length(gp)
xi = gp(ig);
iw = gw(ig);
Se = Se + dphi(xi)'*dphi(xi)*1/h*1*iw;
Me = Me + phi(xi)'*phi(xi)*h*1*iw;
x = phi(xi)*xc';
fe = fe + phi(xi)' * f(x) * h * 1 * iw;
end
% Assembly
S(inod,inod) = S(inod, inod) + Se;
M(inod,inod) = M(inod, inod) + Me;
F(inod) = F(inod) + fe;
end
S = alpha*S + beta*M;
g = zeros(neq,1);
g(1) = -alpha*bn1;
g(end) = alpha*bn2;
alldofs = 1:neq;
u = zeros(neq,1); %Pre-allocate
F = F + g;
u(alldofs) = S(alldofs,alldofs)\F(alldofs)
Warning: Matrix is singular to working precision.
u = 8×1
NaN
NaN
NaN
NaN
NaN
NaN
NaN
NaN
figure
fplot(ue,[a,b], 'r', 'LineWidth',2)
hold on
plot(xnod, u, 'bo')
for iel = 1:N
inod = nodes(iel,:);
xc = xnod(inod);
U = u(inod);
xi = linspace(0,1,100)';
Ue = phi(xi)*U;
Xe = phi(xi)*xc';
plot(Xe,Ue,'b -')
end
% Gauss function for calculate the integral
function [x, w, A] = gauss(n, a, b)
n = 1:(n - 1);
beta = 1 ./ sqrt(4 - 1 ./ (n .* n));
J = diag(beta, 1) + diag(beta, -1);
[V, D] = eig(J);
x = diag(D);
A = b - a;
w = V(1, :) .* V(1, :);
w = w';
x=x';
end
You can find the same post under MATLAB site for syntax highlighting.
Thanks
I tried to read courses, search in different documentation and modify my code without success.
How can I vectorize this code? At the moment it runs so slow. I'm really stuck and I've spent the last couple of hours trying to vectorize it, however I can't seem to get it to work correctly.
My naive program below works incredibly slowly. N should really be 10,000 but the program is struggling with N = 100. Any advice would be appreciated.
The code wants to iterate through the functions given N times for each value w21. It then plots the last 200 values for each value of w21. The code below does work as expected in terms of the plot but as mentioned is far to slow since for a good plot the values need to be in the thousands.
hold on
% Number of iterations
N = 100;
x = 1;
y = 1;
z = 1;
for w21 = linspace(-12,-3,N)
for i = 1:N-1
y = y_iterate(x,z,w21);
z = z_iterate(y);
x = x_iterate(y);
if i >= (N - 200)
p = plot(w21,x,'.k','MarkerSize',3);
end
end
end
Required functions:
function val = x_iterate(y)
val = -3 + 8.*(1 ./ (1 + exp(-y)));
end
function val = z_iterate(y)
val = -7 + 8.*(1 ./ (1 + exp(-y)));
end
function val = y_iterate(x,z,w21)
val = 4 + w21.*(1 ./ (1 + exp(-x))) + 6.*(1 ./ (1 + exp(-z)));
end
I believe it's because of plot. Try:
[X,Y,Z] = deal( zeros(N,N-1) );
w21 = linspace(-12,-3,N);
for i = 1:N
for j = 1:N-1
y = y_iterate(x,z,w21(i));
z = z_iterate(y);
x = x_iterate(y);
X(i,j) = x;
Y(i,j) = y;
Z(i,j) = z;
end
end
nn = max(1,N-200);
plot(w21,X(nn:end,:),'.k')
I asked a question a few days before but I guess it was a little too complicated and I don't expect to get any answer.
My problem is that I need to use ANN for classification. I've read that much better cost function (or loss function as some books specify) is the cross-entropy, that is J(w) = -1/m * sum_i( yi*ln(hw(xi)) + (1-yi)*ln(1 - hw(xi)) ); i indicates the no. data from training matrix X. I tried to apply it in MATLAB but I find it really difficult. There are couple things I don't know:
should I sum each outputs given all training data (i = 1, ... N, where N is number of inputs for training)
is the gradient calculated correctly
is the numerical gradient (gradAapprox) calculated correctly.
I have following MATLAB codes. I realise I may ask for trivial thing but anyway I hope someone can give me some clues how to find the problem. I suspect the problem is to calculate gradients.
Many thanks.
Main script:
close all
clear all
L = #(x) (1 + exp(-x)).^(-1);
NN = #(x,theta) theta{2}*[ones(1,size(x,1));L(theta{1}*[ones(size(x,1),1) x]')];
% theta = [10 -30 -30];
x = [0 0; 0 1; 1 0; 1 1];
y = [0.9 0.1 0.1 0.1]';
theta0 = 2*rand(9,1)-1;
options = optimset('gradObj','on','Display','iter');
thetaVec = fminunc(#costFunction,theta0,options,x,y);
theta = cell(2,1);
theta{1} = reshape(thetaVec(1:6),[2 3]);
theta{2} = reshape(thetaVec(7:9),[1 3]);
NN(x,theta)'
Cost function:
function [jVal,gradVal,gradApprox] = costFunction(thetaVec,x,y)
persistent index;
% 1 x x
% 1 x x
% 1 x x
% x = 1 x x
% 1 x x
% 1 x x
% 1 x x
m = size(x,1);
if isempty(index) || index > size(x,1)
index = 1;
end
L = #(x) (1 + exp(-x)).^(-1);
NN = #(x,theta) theta{2}*[ones(1,size(x,1));L(theta{1}*[ones(size(x,1),1) x]')];
theta = cell(2,1);
theta{1} = reshape(thetaVec(1:6),[2 3]);
theta{2} = reshape(thetaVec(7:9),[1 3]);
Dew = cell(2,1);
DewApprox = cell(2,1);
% Forward propagation
a0 = x(index,:)';
z1 = theta{1}*[1;a0];
a1 = L(z1);
z2 = theta{2}*[1;a1];
a2 = L(z2);
% Back propagation
d2 = 1/m*(a2 - y(index))*L(z2)*(1-L(z2));
Dew{2} = [1;a1]*d2;
d1 = [1;a1].*(1 - [1;a1]).*theta{2}'*d2;
Dew{1} = [1;a0]*d1(2:end)';
% NNRes = NN(x,theta)';
% jVal = -1/m*sum(NNRes-y)*NNRes*(1-NNRes);
jVal = -1/m*(a2 - y(index))*a2*(1-a2);
gradVal = [Dew{1}(:);Dew{2}(:)];
gradApprox = CalcGradApprox(0.0001);
index = index + 1;
function output = CalcGradApprox(epsilon)
output = zeros(size(gradVal));
for n=1:length(thetaVec)
thetaVecMin = thetaVec;
thetaVecMax = thetaVec;
thetaVecMin(n) = thetaVec(n) - epsilon;
thetaVecMax(n) = thetaVec(n) + epsilon;
thetaMin = cell(2,1);
thetaMax = cell(2,1);
thetaMin{1} = reshape(thetaVecMin(1:6),[2 3]);
thetaMin{2} = reshape(thetaVecMin(7:9),[1 3]);
thetaMax{1} = reshape(thetaVecMax(1:6),[2 3]);
thetaMax{2} = reshape(thetaVecMax(7:9),[1 3]);
a2min = NN(x(index,:),thetaMin)';
a2max = NN(x(index,:),thetaMax)';
jValMin = -1/m*(a2min-y(index))*a2min*(1-a2min);
jValMax = -1/m*(a2max-y(index))*a2max*(1-a2max);
output(n) = (jValMax - jValMin)/2/epsilon;
end
end
end
EDIT:
Below I present the correct version of my costFunction for those who may be interested.
function [jVal,gradVal,gradApprox] = costFunction(thetaVec,x,y)
m = size(x,1);
L = #(x) (1 + exp(-x)).^(-1);
NN = #(x,theta) L(theta{2}*[ones(1,size(x,1));L(theta{1}*[ones(size(x,1),1) x]')]);
theta = cell(2,1);
theta{1} = reshape(thetaVec(1:6),[2 3]);
theta{2} = reshape(thetaVec(7:9),[1 3]);
Delta = cell(2,1);
Delta{1} = zeros(size(theta{1}));
Delta{2} = zeros(size(theta{2}));
D = cell(2,1);
D{1} = zeros(size(theta{1}));
D{2} = zeros(size(theta{2}));
jVal = 0;
for in = 1:size(x,1)
% Forward propagation
a1 = [1;x(in,:)']; % added bias to a0
z2 = theta{1}*a1;
a2 = [1;L(z2)]; % added bias to a1
z3 = theta{2}*a2;
a3 = L(z3);
% Back propagation
d3 = a3 - y(in);
d2 = theta{2}'*d3.*a2.*(1 - a2);
Delta{2} = Delta{2} + d3*a2';
Delta{1} = Delta{1} + d2(2:end)*a1';
jVal = jVal + sum( y(in)*log(a3) + (1-y(in))*log(1-a3) );
end
D{1} = 1/m*Delta{1};
D{2} = 1/m*Delta{2};
jVal = -1/m*jVal;
gradVal = [D{1}(:);D{2}(:)];
gradApprox = CalcGradApprox(x(in,:),0.0001);
% Nested function to calculate gradApprox
function output = CalcGradApprox(x,epsilon)
output = zeros(size(thetaVec));
for n=1:length(thetaVec)
thetaVecMin = thetaVec;
thetaVecMax = thetaVec;
thetaVecMin(n) = thetaVec(n) - epsilon;
thetaVecMax(n) = thetaVec(n) + epsilon;
thetaMin = cell(2,1);
thetaMax = cell(2,1);
thetaMin{1} = reshape(thetaVecMin(1:6),[2 3]);
thetaMin{2} = reshape(thetaVecMin(7:9),[1 3]);
thetaMax{1} = reshape(thetaVecMax(1:6),[2 3]);
thetaMax{2} = reshape(thetaVecMax(7:9),[1 3]);
a3min = NN(x,thetaMin)';
a3max = NN(x,thetaMax)';
jValMin = 0;
jValMax = 0;
for inn=1:size(x,1)
jValMin = jValMin + sum( y(inn)*log(a3min) + (1-y(inn))*log(1-a3min) );
jValMax = jValMax + sum( y(inn)*log(a3max) + (1-y(inn))*log(1-a3max) );
end
jValMin = 1/m*jValMin;
jValMax = 1/m*jValMax;
output(n) = (jValMax - jValMin)/2/epsilon;
end
end
end
I've only had a quick eyeball over your code. Here are some pointers.
Q1
should I sum each outputs given all training data (i = 1, ... N, where
N is number of inputs for training)
If you are talking in relation to the cost function, it is normal to sum and normalise by the number of training examples in order to provide comparison between.
I can't tell from the code whether you have a vectorised implementation which will change the answer. Note that the sum function will only sum up a single dimension at a time - meaning if you have a (M by N) array, sum will result in a 1 by N array.
The cost function should have a scalar output.
Q2
is the gradient calculated correctly
The gradient is not calculated correctly - specifically the deltas look wrong. Try following Andrew Ng's notes [PDF] they are very good.
Q3
is the numerical gradient (gradAapprox) calculated correctly.
This line looks a bit suspect. Does this make more sense?
output(n) = (jValMax - jValMin)/(2*epsilon);
EDIT: I actually can't make heads or tails of your gradient approximation. You should only use forward propagation and small tweaks in the parameters to compute the gradient. Good luck!
I've written some code to implement an algorithm that takes as input a vector q of real numbers, and returns as an output a complex matrix R. The Matlab code below produces a plot showing the input vector q and the output matrix R.
Given only the complex matrix output R, I would like to obtain the input vector q. Can I do this using least-squares optimization? Since there is a recursive running sum in the code (rs_r and rs_i), the calculation for a column of the output matrix is dependent on the calculation of the previous column.
Perhaps a non-linear optimization can be set up to recompose the input vector q from the output matrix R?
Looking at this in another way, I've used an algorithm to compute a matrix R. I want to run the algorithm "in reverse," to get the input vector q from the output matrix R.
If there is no way to recompose the starting values from the output, thereby treating the problem as a "black box," then perhaps the mathematics of the model itself can be used in the optimization? The program evaluates the following equation:
The Utilde(tau, omega) is the output matrix R. The tau (time) variable comprises the columns of the response matrix R, whereas the omega (frequency) variable comprises the rows of the response matrix R. The integration is performed as a recursive running sum from tau = 0 up to the current tau timestep.
Here are the plots created by the program posted below:
Here is the full program code:
N = 1001;
q = zeros(N, 1); % here is the input
q(1:200) = 55;
q(201:300) = 120;
q(301:400) = 70;
q(401:600) = 40;
q(601:800) = 100;
q(801:1001) = 70;
dt = 0.0042;
fs = 1 / dt;
wSize = 101;
Glim = 20;
ginv = 0;
R = get_response(N, q, dt, wSize, Glim, ginv); % R is output matrix
rows = wSize;
cols = N;
figure; plot(q); title('q value input as vector');
ylim([0 200]); xlim([0 1001])
figure; imagesc(abs(R)); title('Matrix output of algorithm')
colorbar
Here is the function that performs the calculation:
function response = get_response(N, Q, dt, wSize, Glim, ginv)
fs = 1 / dt;
Npad = wSize - 1;
N1 = wSize + Npad;
N2 = floor(N1 / 2 + 1);
f = (fs/2)*linspace(0,1,N2);
omega = 2 * pi .* f';
omegah = 2 * pi * f(end);
sigma2 = exp(-(0.23*Glim + 1.63));
sign = 1;
if(ginv == 1)
sign = -1;
end
ratio = omega ./ omegah;
rs_r = zeros(N2, 1);
rs_i = zeros(N2, 1);
termr = zeros(N2, 1);
termi = zeros(N2, 1);
termr_sub1 = zeros(N2, 1);
termi_sub1 = zeros(N2, 1);
response = zeros(N2, N);
% cycle over cols of matrix
for ti = 1:N
term0 = omega ./ (2 .* Q(ti));
gamma = 1 / (pi * Q(ti));
% calculate for the real part
if(ti == 1)
Lambda = ones(N2, 1);
termr_sub1(1) = 0;
termr_sub1(2:end) = term0(2:end) .* (ratio(2:end).^-gamma);
else
termr(1) = 0;
termr(2:end) = term0(2:end) .* (ratio(2:end).^-gamma);
rs_r = rs_r - dt.*(termr + termr_sub1);
termr_sub1 = termr;
Beta = exp( -1 .* -0.5 .* rs_r );
Lambda = (Beta + sigma2) ./ (Beta.^2 + sigma2); % vector
end
% calculate for the complex part
if(ginv == 1)
termi(1) = 0;
termi(2:end) = (ratio(2:end).^(sign .* gamma) - 1) .* omega(2:end);
else
termi = (ratio.^(sign .* gamma) - 1) .* omega;
end
rs_i = rs_i - dt.*(termi + termi_sub1);
termi_sub1 = termi;
integrand = exp( 1i .* -0.5 .* rs_i );
if(ginv == 1)
response(:,ti) = Lambda .* integrand;
else
response(:,ti) = (1 ./ Lambda) .* integrand;
end
end % ti loop
No, you cannot do so unless you know the "model" itself for this process. If you intend to treat the process as a complete black box, then it is impossible in general, although in any specific instance, anything can happen.
Even if you DO know the underlying process, then it may still not work, as any least squares estimator is dependent on the starting values, so if you do not have a good guess there, it may converge to the wrong set of parameters.
It turns out that by using the mathematics of the model, the input can be estimated. This is not true in general, but for my problem it seems to work. The cumulative integral is eliminated by a partial derivative.
N = 1001;
q = zeros(N, 1);
q(1:200) = 55;
q(201:300) = 120;
q(301:400) = 70;
q(401:600) = 40;
q(601:800) = 100;
q(801:1001) = 70;
dt = 0.0042;
fs = 1 / dt;
wSize = 101;
Glim = 20;
ginv = 0;
R = get_response(N, q, dt, wSize, Glim, ginv);
rows = wSize;
cols = N;
cut_val = 200;
imagLogR = imag(log(R));
Mderiv = zeros(rows, cols-2);
for k = 1:rows
val = deriv_3pt(imagLogR(k,:), dt);
val(val > cut_val) = 0;
Mderiv(k,:) = val(1:end-1);
end
disp('Running iteration');
q0 = 10;
q1 = 500;
NN = cols - 2;
qout = zeros(NN, 1);
for k = 1:NN
data = Mderiv(:,k);
qout(k) = fminbnd(#(q) curve_fit_to_get_q(q, dt, rows, data),q0,q1);
end
figure; plot(q); title('q value input as vector');
ylim([0 200]); xlim([0 1001])
figure;
plot(qout); title('Reconstructed q')
ylim([0 200]); xlim([0 1001])
Here are the supporting functions:
function output = deriv_3pt(x, dt)
% Function to compute dx/dt using the 3pt symmetrical rule
% dt is the timestep
N = length(x);
N0 = N - 1;
output = zeros(N0, 1);
denom = 2 * dt;
for k = 2:N0
output(k - 1) = (x(k+1) - x(k-1)) / denom;
end
function sse = curve_fit_to_get_q(q, dt, rows, data)
fs = 1 / dt;
N2 = rows;
f = (fs/2)*linspace(0,1,N2); % vector for frequency along cols
omega = 2 * pi .* f';
omegah = 2 * pi * f(end);
ratio = omega ./ omegah;
gamma = 1 / (pi * q);
termi = ((ratio.^(gamma)) - 1) .* omega;
Error_Vector = termi - data;
sse = sum(Error_Vector.^2);
I'm trying to write a cubic spline interpolation program. I have written the program but, the graph is not coming out correctly. The spline uses natural boundary conditions(second dervative at start/end node are 0). The code is in Matlab and is shown below,
clear all
%Function to Interpolate
k = 10; %Number of Support Nodes-1
xs(1) = -1;
for j = 1:k
xs(j+1) = -1 +2*j/k; %Support Nodes(Equidistant)
end;
fs = 1./(25.*xs.^2+1); %Support Ordinates
x = [-0.99:2/(2*k):0.99]; %Places to Evaluate Function
fx = 1./(25.*x.^2+1); %Function Evaluated at x
%Cubic Spline Code(Coefficients to Calculate 2nd Derivatives)
f(1) = 2*(xs(3)-xs(1));
g(1) = xs(3)-xs(2);
r(1) = (6/(xs(3)-xs(2)))*(fs(3)-fs(2)) + (6/(xs(2)-xs(1)))*(fs(1)-fs(2));
e(1) = 0;
for i = 2:k-2
e(i) = xs(i+1)-xs(i);
f(i) = 2*(xs(i+2)-xs(i));
g(i) = xs(i+2)-xs(i+1);
r(i) = (6/(xs(i+2)-xs(i+1)))*(fs(i+2)-fs(i+1)) + ...
(6/(xs(i+1)-xs(i)))*(fs(i)-fs(i+1));
end
e(k-1) = xs(k)-xs(k-1);
f(k-1) = 2*(xs(k+1)-xs(k-1));
r(k-1) = (6/(xs(k+1)-xs(k)))*(fs(k+1)-fs(k)) + ...
(6/(xs(k)-xs(k-1)))*(fs(k-1)-fs(k));
%Tridiagonal System
i = 1;
A = zeros(k-1,k-1);
while i < size(A)+1;
A(i,i) = f(i);
if i < size(A);
A(i,i+1) = g(i);
A(i+1,i) = e(i);
end
i = i+1;
end
for i = 2:k-1 %Decomposition
e(i) = e(i)/f(i-1);
f(i) = f(i)-e(i)*g(i-1);
end
for i = 2:k-1 %Forward Substitution
r(i) = r(i)-e(i)*r(i-1);
end
xn(k-1)= r(k-1)/f(k-1);
for i = k-2:-1:1 %Back Substitution
xn(i) = (r(i)-g(i)*xn(i+1))/f(i);
end
%Interpolation
if (max(xs) <= max(x))
error('Outside Range');
end
if (min(xs) >= min(x))
error('Outside Range');
end
P = zeros(size(length(x),length(x)));
i = 1;
for Counter = 1:length(x)
for j = 1:k-1
a(j) = x(Counter)- xs(j);
end
i = find(a == min(a(a>=0)));
if i == 1
c1 = 0;
c2 = xn(1)/6/(xs(2)-xs(1));
c3 = fs(1)/(xs(2)-xs(1));
c4 = fs(2)/(xs(2)-xs(1))-xn(1)*(xs(2)-xs(1))/6;
t1 = c1*(xs(2)-x(Counter))^3;
t2 = c2*(x(Counter)-xs(1))^3;
t3 = c3*(xs(2)-x(Counter));
t4 = c4*(x(Counter)-xs(1));
P(Counter) = t1 +t2 +t3 +t4;
else
if i < k-1
c1 = xn(i-1+1)/6/(xs(i+1)-xs(i-1+1));
c2 = xn(i+1)/6/(xs(i+1)-xs(i-1+1));
c3 = fs(i-1+1)/(xs(i+1)-xs(i-1+1))-xn(i-1+1)*(xs(i+1)-xs(i-1+1))/6;
c4 = fs(i+1)/(xs(i+1)-xs(i-1+1))-xn(i+1)*(xs(i+1)-xs(i-1+1))/6;
t1 = c1*(xs(i+1)-x(Counter))^3;
t2 = c2*(x(Counter)-xs(i-1+1))^3;
t3 = c3*(xs(i+1)-x(Counter));
t4 = c4*(x(Counter)-xs(i-1+1));
P(Counter) = t1 +t2 +t3 +t4;
else
c1 = xn(i-1+1)/6/(xs(i+1)-xs(i-1+1));
c2 = 0;
c3 = fs(i-1+1)/(xs(i+1)-xs(i-1+1))-xn(i-1+1)*(xs(i+1)-xs(i-1+1))/6;
c4 = fs(i+1)/(xs(i+1)-xs(i-1+1));
t1 = c1*(xs(i+1)-x(Counter))^3;
t2 = c2*(x(Counter)-xs(i-1+1))^3;
t3 = c3*(xs(i+1)-x(Counter));
t4 = c4*(x(Counter)-xs(i-1+1));
P(Counter) = t1 +t2 +t3 +t4;
end
end
end
P = P';
P(length(x)) = NaN;
plot(x,P,x,fx)
When I run the code, the interpolation function is not symmetric and, it doesn't converge correctly. Can anyone offer any suggestions about problems in my code? Thanks.
I wrote a cubic spline package in Mathematica a long time ago. Here is my translation of that package into Matlab. Note I haven't looked at cubic splines in about 7 years, so I'm basing this off my own documentation. You should check everything I say.
The basic problem is we are given n data points (x(1), y(1)) , ... , (x(n), y(n)) and we wish to calculate a piecewise cubic interpolant. The interpolant is defined as
S(x) = { Sk(x) when x(k) <= x <= x(k+1)
{ 0 otherwise
Here Sk(x) is a cubic polynomial of the form
Sk(x) = sk0 + sk1*(x-x(k)) + sk2*(x-x(k))^2 + sk3*(x-x(k))^3
The properties of the spline are:
The spline pass through the data point Sk(x(k)) = y(k)
The spline is continuous at the end-points and thus continuous everywhere in the interpolation interval Sk(x(k+1)) = Sk+1(x(k+1))
The spline has continuous first derivative Sk'(x(k+1)) = Sk+1'(x(k+1))
The spline has continuous second derivative Sk''(x(k+1)) = Sk+1''(x(k+1))
To construct a cubic spline from a set of data point we need to solve for the coefficients
sk0, sk1, sk2 and sk3 for each of the n-1 cubic polynomials. That is a total of 4*(n-1) = 4*n - 4 unknowns. Property 1 supplies n constraints, and properties 2,3,4 each supply an additional n-2 constraints. Thus we have n + 3*(n-2) = 4*n - 6 constraints and 4*n - 4 unknowns. This leaves two degrees of freedom. We fix these degrees of freedom by setting the second derivative equal to zero at the start and end nodes.
Let m(k) = Sk''(x(k)) , h(k) = x(k+1) - x(k) and d(k) = (y(k+1) - y(k))/h(k). The following
three-term recurrence relation holds
h(k-1)*m(k-1) + 2*(h(k-1) + h(k))*m(k) + h(k)*m(k+1) = 6*(d(k) - d(k-1))
The m(k) are unknowns we wish to solve for. The h(k) and d(k) are defined by the input data.
This three-term recurrence relation defines a tridiagonal linear system. Once the m(k) are determined the coefficients for Sk are given by
sk0 = y(k)
sk1 = d(k) - h(k)*(2*m(k) + m(k-1))/6
sk2 = m(k)/2
sk3 = m(k+1) - m(k)/(6*h(k))
Okay that is all the math you need to know to completely define the algorithm to compute a cubic spline. Here it is in Matlab:
function [s0,s1,s2,s3]=cubic_spline(x,y)
if any(size(x) ~= size(y)) || size(x,2) ~= 1
error('inputs x and y must be column vectors of equal length');
end
n = length(x)
h = x(2:n) - x(1:n-1);
d = (y(2:n) - y(1:n-1))./h;
lower = h(1:end-1);
main = 2*(h(1:end-1) + h(2:end));
upper = h(2:end);
T = spdiags([lower main upper], [-1 0 1], n-2, n-2);
rhs = 6*(d(2:end)-d(1:end-1));
m = T\rhs;
% Use natural boundary conditions where second derivative
% is zero at the endpoints
m = [ 0; m; 0];
s0 = y;
s1 = d - h.*(2*m(1:end-1) + m(2:end))/6;
s2 = m/2;
s3 =(m(2:end)-m(1:end-1))./(6*h);
Here is some code to plot a cubic spline:
function plot_cubic_spline(x,s0,s1,s2,s3)
n = length(x);
inner_points = 20;
for i=1:n-1
xx = linspace(x(i),x(i+1),inner_points);
xi = repmat(x(i),1,inner_points);
yy = s0(i) + s1(i)*(xx-xi) + ...
s2(i)*(xx-xi).^2 + s3(i)*(xx - xi).^3;
plot(xx,yy,'b')
plot(x(i),0,'r');
end
Here is a function that constructs a cubic spline and plots in on the famous Runge function:
function cubic_driver(num_points)
runge = #(x) 1./(1+ 25*x.^2);
x = linspace(-1,1,num_points);
y = runge(x);
[s0,s1,s2,s3] = cubic_spline(x',y');
plot_points = 1000;
xx = linspace(-1,1,plot_points);
yy = runge(xx);
plot(xx,yy,'g');
hold on;
plot_cubic_spline(x,s0,s1,s2,s3);
You can see it in action by running the following at the Matlab prompt
>> cubic_driver(5)
>> clf
>> cubic_driver(10)
>> clf
>> cubic_driver(20)
By the time you have twenty nodes your interpolant is visually indistinguishable from the Runge function.
Some comments on the Matlab code: I don't use any for or while loops. I am able to vectorize all operations. I quickly form the sparse tridiagonal matrix with spdiags. I solve it using the backslash operator. I counting on Tim Davis's UMFPACK to handle the decomposition and forward and backward solves.
Hope that helps. The code is available as a gist on github https://gist.github.com/1269709
There was a bug in spline function, generated (n-2) by (n-2) should be symmetric:
lower = h(2:end);
main = 2*(h(1:end-1) + h(2:end));
upper = h(1:end-1);
http://www.mpi-hd.mpg.de/astrophysik/HEA/internal/Numerical_Recipes/f3-3.pdf