Sparse Matirx

How to code sparse matrix from the existing matrix?

Answer 1

[Andrei Bobrov]

A = rand(10)

A(A>.15) = 0 % existing matrix
Asp = sparse(A) % sparse matrix

[KAKOLI DEY]

if I take temperature data matrix how would i convert yhis into a sparse

Answer 2

[John D'Errico]

From what existing matrix? In what form?

If you already have a matrix, then the sparse function will convert it to sparse form. Thus

A = rand(6);
A(A > .1) = 0;

A here is a FULL matrix, even though it is mostly zero.

A
A =
          0            0            0            0            0            0
          0     0.084436            0            0            0            0
          0            0            0            0     0.075967     0.049654
  0.0046342            0            0            0            0            0
          0            0            0            0            0            0
          0            0            0            0            0            0

As you can see, A is full, just a double precision matrix.

whos A
Name      Size            Bytes  Class     Attributes
A         6x6               288  double

A is convertible into sparse form, trivially, using sparse.

As = sparse(A)
As =
   (4,1)       0.0046342
   (2,2)        0.084436
   (3,5)        0.075967
   (3,6)        0.049654

whos As
  Name      Size            Bytes  Class     Attributes
  As        6x6               120  double    sparse

If you have a matrix in some other form, perhaps a list of row and column indices and values, then again, sparse does the job.

Or, perhaps you have the matrix in the form of known diagonals. Then spdiags is the proper tool.

But we cannot read your mind as to what you have. Just telling us that you have a "temperature data matrix" is nothing of value. It is insufficient information.