storage format using SPARSE and INT8

Question

Giovanni Gardan 2020-5-27

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/535343-storage-format-using-sparse-and-int8

I'd like to save storage and time when I run my code. In order to do this I'm trying to modify the format-storage of some (1 0 -1) matrixes and in the follow example I note that int8 is the best. That's ok!

But if I try to save B as a sparse matrix (see the comment) I get the following error:

Undefined function 'sparse' for input arguments of type 'int8'.

1) So, is it possible to save a sparse matrix (made only with the values of element 1, 0 or -1) not with double precision, but with int storage-format to save too much storage?

2) In general, does making operation between different saved elements (for example the sum of matrix B with another matrix, but saved in double precision) create problems??

%Code.m
A = [1 0 0 0 0 0 0 0 0 0 0 0 0;
    -1 0 0 0 0 0 0 0 0 0 0 0 1;
     0 0 0 0 0 0 0 0 0 1 0 0 0];
B= int8(A); 
C = sparse(A);
% D=sparse(B)           <<<<<<<<<HERE THE COMMENT
whos A B C
-------------------------------------------------------------------------------------------------
%Command window shows
>> Code
  Name      Size            Bytes  Class     Attributes
  A         3x13              312  double              
  B         3x13               39  int8                
  C         3x13              176  double    sparse   