Integer check

How can i do an integer check in matlab? Say i have a number, a = 1.5, how do i check if it is integer or not? i want to use an if statement:

if non_integer(a) == true

end

so i need to know what to use for "non_integer", which would return a true or false

[Phani]

I think you can use rem function which throws out the remainder.

if(rem(a,1) ~=0) %checks if a is an integer with help of remainder

....

end

Answer 1

[Sean de Wolski]

floor(x)==x

[Walter Roberson]

Fails on inf and -inf ;-)

[Jukka Kaipala]

Like it should. Inf is not considered an integer, and the same goes for NaN.

[Thorsten]

Like it should not. Because Inf is not an integer, but

floor(inf) == inf

returns true.

So the correct test would be

~isinf(x) & floor(x) == x

Or, as Walter suggests,

mod(x,1) == 0

But this may be all just of theoretical interest.

Answer 2

[Walter Roberson]

I suggest you think about floor(), ceil(), fix(), and round()

[Walter Roberson]

A test that works even for nan and inf and -inf :

mod(x,1) == 0

[Walter Roberson]

NaN == 0 is false, so in each of those cases, the result of the test would be false, indicating that the values are not integers.

But if you tried the one above, floor(inf)==inf then the result would be true, seemingly indicating that it is an integer when it is not.

Answer 3

[Fandi Bataineh]

use

floor(x)==ceil(x)

this will be true only for integers

[Sivyer Ge]

Actually this will not work for type 'char'. Do you know other way to identify the difference between char and integer?

Answer 4

[Manuel Alcazar]

function [bool,idx] = isint(x)
    % Check whether input is integer or not
    % Inf and NaN are not integers
    if ~isnumeric(x)
       error('Input must be a numeric, not a %s.',class(x))
    end
    bool = (mod(x,1) == 0);
%     bool = (round(x) == floor(x));  % Other approach. Fails with Inf and
%     NaN.
    idx = find(bool); 
% Manolín Sept-2019
end

Answer 5

[per isakson]

function    isf = isflint( m )

%   floating double only
%   
%   http://www.mathworks.com/company/newsletters/news_notes/pdf/Fall96Cleve.pdf
%   JanSimon: http://www.mathworks.se/matlabcentral/answers/67247-about-isinteger- ...
%       command-confusion#answer_78914   

    assert( isa( m, 'double' ) && isvector( m )     ...
        ,   'isflint:IllegalInput'                  ...
        ,   'The input should be a double vector'   )

    isf =   all( abs(   m ) <= flintmax ) ...
        &&  all( floor( m ) == m        ) ; 
end

[Stephen23]

This could be trivially vectorized: why force it to return a scalar?