Matlab实现向前欧拉方法遇下标错误求助：下标需为正整数或逻辑值

Hey there, let's break down what's causing this error and fix your Forward Euler implementation step by step!

The Root of the Problem #

MATLAB uses 1-based indexing for arrays—meaning the first element of any array is always at index 1, not 0. Your code tries to access t(0) and y(0), which are invalid indices because 0 isn't a positive integer. That's exactly why you're getting that error on line 5 (and you'd hit the same issue on the y(0)=y0 line too).

Corrected Code #

Here's the fixed version of your function, with comments explaining the changes:

function y = ForwardEulerMethod(f,y0,T,N)
    h = T/N;
    t = zeros(N+1,1);
    % Loop from 1 to N+1 (valid indices) instead of 0 to N
    for i = 1:N+1
        t(i) = (i-1)*h;  % Shift to get t(1)=0, t(2)=h, ..., t(N+1)=T
    end
    y = zeros(N+1,1);
    y(1) = y0;  % Use index 1 for the initial condition
    % Loop from 2 to N+1 to compute subsequent values
    for i = 2:N+1
        y(i) = y(i-1) + h*f(t(i-1), y(i-1));
    end
end

Key Changes Explained #

• First loop adjustment: We start i at 1 instead of 0, and calculate t(i) as (i-1)*h. This gives us the correct time points: t(1) = 0, t(2) = h, all the way to t(N+1) = T.
• Initial condition index: We assign y0 to y(1) instead of y(0)—this is the valid first element of the y array.
• Second loop adjustment: We start i at 2 (since y(1) is already set) and go up to N+1, ensuring we always access valid indices i-1 (which starts at 1) and i.

Quick Test to Verify #

You can test the function with a simple ODE like dy/dt = y (solution y(t) = e^t) to make sure it works:

f = @(t,y) y;
y0 = 1;
T = 1;
N = 10;
y = ForwardEulerMethod(f,y0,T,N);
% Compare with analytical solution
t = linspace(0,T,N+1);
y_analytical = exp(t);
plot(t,y,'o-',t,y_analytical,'--');
legend('Forward Euler','Analytical Solution');

This should plot the Euler approximation alongside the exact solution, confirming the code runs without errors.

内容的提问来源于stack exchange，提问作者Matt Robbins