Scilab Home page | Wiki | Bug tracker | Forge | Mailing list archives | ATOMS | File exchange
Please login or create an account
Change language to: English - Português - 日本語 - Русский
Aide de Scilab >> Scilab > Mots clés Scilab > comparison

comparison

comparison, relational operators

Syntax

a==b
a~=b or a<>b
a<b
a<=b
a>b
a>=b

Arguments

a

any type of variable for a==b, a~=b a<>b equality comparisons and restricted to real floating point and integer array for order related comparisons a<b, a<=b, a>b, a>=b.

b

any type of variable for a==b, a~=b a<>b equality comparisons and restricted to real floating point and integer arrays for order related comparisons a<b, a<=b, a>b, a>=b.

Description

Two classes of operators have to be distinguished:

The equality and inequality comparisons:

a==b, a~=b (or equivalently a<>b). These operators apply to any type of operands.

The order related comparisons:

a<b, a<=b, a>b, a>=b. These operators apply only to integer and real numbers.

The result of the comparison operators also depends on the operands types:

With array variables

arrays of numbers, boolean arrays, string arrays, polynomial and rational arrays, handle arrays, lists... the following rules apply:

  • If a and b are arrays with same types and sizes, the comparison is performed element by element and the result is an array of booleans of the same sizes.

  • If a and b have the same type and a or b is a scalar, then the scalar is compared with each element of the other array. The result is an array of booleans of the size of the non scalar operand.

  • If a or b is an array while the other one is empty, then the result is a scalar boolean.

  • In all other cases, the result is %F

  • If the operand data types are different but "compatible" like decimal numbers and encoded integers, then a type conversion is performed before the comparison.

With other type of operands

like function, libraries, the result is %t if the objects are identical and %f in the other cases.

Equality comparison between operands of incompatible data types returns %f.

  • Polynomials: %s==%z returns %F : Two polynomials can't be considered as equal if they are not about the same variable.

  • Rationals: 1/%s == 1/%z returns %F : Two rationals can't be equal if they are not about the same variable. In addition, the same rational may have different unnormalized forms that won't be considered as equal. For instance, 2/(2*%z) == 1/%z returns %F, even in simp_mode(%T) simplification mode.

  • Sparse numerical matrix: any element-wise comparison with a full-encoded numerical array always returns a boolean sparse array. See examples.

  • Graphics identifiers: A graphic handle h1 and its copy h2=h1 will always stay equal, even if the graphics is changed through one of them. See examples.

Examples

//element wise comparisons
(1:5)==3
(1:5)<=4
(1:5)<=[1 4 2 3 0]
1<[]
list(1,2,3)~=list(1,3,3)
"foo"=="bar"
sparse([1,2;4,5;3,10],[1,2,3]) == sparse([1,2;4,5;3,10],[1,2,3])

//object wise comparisons
(1:10)==[4,3]
'foo'==3
1==[]
list(1,2,3)==1

isequal(list(1,2,3),1)
isequal(1:10,1)

Comparisons with implicit conversion of type or encoding:

int32(1) == 1
int32(1) < 1.5
int32(1:5) < int8(3)
1 == complex(1,0)
1 > complex(1,0)  // still OK, but..
1 > complex(1,1)  // => error: complex numbers not orderable
--> int32(1) == 1
 ans  =
  T

--> int32(1) < 1.5
 ans  =
  T

--> int32(1:5) < int8(3)
 ans  =
  T T F F F

--> 1 == complex(1,0)
 ans  =
  T

--> 1 > complex(1,0)  // still OK, but..
 ans  =
  F

--> 1 > complex(1,1)  // => error: complex numbers not orderable
at line    11 of function %s_2_s ( SCI\modules\overloading\macros\%s_2_s.sci line 23 )
Complex comparison not supported. Please define %s_2_s_custom() or check your code.

Comparisons with polynomials and rationals:

p = 0*%s
p == 0
r = p/(1+0*%s)
r == 0
r == p
ps = (1-%s)^2, pz = (1-%z)^2
ps == pz  // => %F : same variable required
--> p = 0*%s
 p  =
   0

--> p == 0
 ans  =
  T

--> r = p/(1+0*%s)
 r  =
   0
   --
   1

--> r == 0
 ans  =
  T

--> r == p
 ans  =
  T

--> ps = (1-%s)^2, pz = (1-%z)^2
 ps  =
             2
   1  -2s +s
 pz  =
             2
   1  -2z +z

--> ps == pz  // => %F : same variable required
 ans  =
  F

Comparisons with a sparse numerical matrix: All element-wise comparisons yield a sparse-encoded result, %F otherwise.

sp = sparse([0 1 0 0 -2 0 4 0 0])
sp < 0
sp == 1
sp >=       [2 3 -1 2 -4 0 3 1 0]
sp == %i
sp == list(3) // => %F
--> sp = sparse([0 1 0 0 -2 0 4 0 0])
 sp  =
(  1,  9) sparse matrix
(  1,  2)     1.
(  1,  5)    -2.
(  1,  7)     4.

--> sp < 0
 ans  =
(  1,  9) sparse matrix
(  1,  5)   T

--> sp == 1
 ans  =
(  1,  9) sparse matrix
(  1,  2)   T

--> sp >= [2 3 -1 2 -4 0 3 1 0]
 ans  =
(  1,  9) sparse matrix
(  1,  3)   T
(  1,  5)   T
(  1,  6)   T
(  1,  7)   T
(  1,  9)   T

--> sp == %i
 ans  =
(  1,  9)False sparse matrix

--> sp == list(3)  // object comparison => dense %F
 ans  =
  F

Comparisons between graphics identifiers:

plot2d()
e1 = gce();
e2 = e1;    // e1 and e2 point to the same graphical object
e2.tag
e1.tag = "3 curves";
e1 == e2
e2.tag
--> e2.tag
 ans  =

--> e1.tag = "3 curves";
--> e1 == e2
 ans  =
  T

--> e2.tag
 ans  =
 3 curves

Comparisons between functions aliases are possible:

sine = sin ;
sine == sin
sinus = sind ;
sinus == sind
sinus(0:90:360)
--> sine = sin ;
--> sine == sin
 ans  =
  T

--> sinus = sind ;
--> sinus == sind
 ans  =
  T

--> sinus(0:90:360)
 ans  =
   0.   1.   0.  -1.   0.

See also

  • less — (<) less than comparison
  • greater
  • equal — (=) assignment , comparison, equal sign
  • isequal — comparison of objects
  • boolean — Objets Scilab, variables booléennes et opérateurs & | ~

History

VersionDescription
6.0
  • ~ (not) priority is now higher than the comparisons one (== ~= < <= >= >).
  • Complex numbers with a null imaginary part are now compared as real numbers.
Scilab Enterprises
Copyright (c) 2011-2017 (Scilab Enterprises)
Copyright (c) 1989-2012 (INRIA)
Copyright (c) 1989-2007 (ENPC)
with contributors
Last updated:
Thu Feb 14 14:59:53 CET 2019