Question

我怎样才能更通用？

XD_L数组是手动介绍。

% Data
Xd_L = [10 100 1000];
C = zeros(length(Xd_L))

% Calc
C12 = -1/Xd_L(1);
C13 = -1/Xd_L(2);
C11 = abs(C12+C13);

C21 = C12;
C23 = -1/Xd_L(3);
C22 = abs(C21+C23);

C31 = C13;
C32 = C23;
C33 = abs(C31+C32);

% Result:
C = [C11 C12 C13; C21 C22 C23; C31 C32 C33]

Answer 1

这是一种通用案例的方法 -

function out = scale_assign(Xd_L)

m = numel(Xd_L);
n = ceil(sqrt(2*m));
off_diag_vals = -1./Xd_L;
[y,x] = find(tril(true(n),-1));

out = zeros(n,n);
out(sub2ind(size(out),y,x)) = off_diag_vals;
out(sub2ind(size(out),x,y)) = off_diag_vals;
out(1:n+1:n^2) = abs(sum(out,2));

样品运行 -

>> Xd_L = [10 100 1000];
>> scale_assign(Xd_L)
ans =
         0.11         -0.1        -0.01
         -0.1        0.101       -0.001
        -0.01       -0.001        0.011
>> Xd_L = [10 100 1000 10000 100000 1000000];
>> scale_assign(Xd_L)
ans =
        0.111         -0.1        -0.01       -0.001
         -0.1      0.10011      -0.0001       -1e-05
        -0.01      -0.0001     0.010101       -1e-06
       -0.001       -1e-05       -1e-06     0.001011

Matlab为矩阵的特定元素赋值

1 个答案: