CODEWITHSUNDEEP
matlablogical-operators
Hugo Trentesaux
Hugo Trentesaux

Reputation: 1999

Why does Matlab returns double as result of logical multiplication?

Is it normal that Matlab returns a double as a result for a logical multiplication ?

A=[true false];
B=[false true];
class(A.*B)

output is 

ans =
double

The best way I found is a cast : logical(A.*B), but I do not feel it is clean.

Upvotes: 1

Views: 514

Answers (2)

Tommaso Belluzzo
Tommaso Belluzzo

Reputation: 23685

In many languages (including Matlab) the boolean type (called logical in Matlab) is not a numeric type, even if it has an underlying numeric type (just like every other type). It's underlying numeric type is nothing but a byte (uint8 in Matlab) whose possible values are constrained to 0 and 1.

In the majority of programming languages, when you try to perform arithmetic operations between booleans, the compiler throws an error telling you that it's not possible. Matlab, on the opposite, performs an implicit conversion of your logicals to double before proceeding with the calculations, and let you go on without errors:

A = [true false]; => A = [1 0];
B = [false true]; => A = [0 1];
ans = A .* B;

whos ans => double

That's why your result type is double.

Actually, if you want to perform operations with logicals, sticking to logical return type... you have to use logical operators. For more details: https://mathworks.com/help/matlab/logical-operations.html

Actually, your result can be achieved staying in the logical domain through:

res = A & B;

Upvotes: 2

Hugo Trentesaux
Hugo Trentesaux

Reputation: 1999

I found ! It is simply A&B

>> A&B
ans =
     0     0

>> class(ans)
ans =
     logical

Upvotes: 0

Related Questions