Note: This page has been translated by MathWorks. Please click here

To view all translated materals including this page, select Japan from the country navigator on the bottom of this page.

To view all translated materals including this page, select Japan from the country navigator on the bottom of this page.

**MuPAD® notebooks are not recommended. Use MATLAB® live scripts instead.**

**MATLAB live scripts support most MuPAD functionality, though there are some differences. For more information, see Convert MuPAD Notebooks to MATLAB Live Scripts.**

If `a`

, `b`

, and `m`

are
integers, and `(a - b)/m`

is also an integer, then
the numbers `a`

and `b`

are congruent
modulo `m`

. The remainder of the division `a/m`

is
equal to the remainder of the division of `b/m`

.
For example, `11 ≡ 5(mod 3)`

:

5 mod 3 = 11 mod 3

For known integers `a`

and `m`

,
all integers `b`

, such that ```
a ≡ b(mod
m)
```

, form a residue class. Thus, the numbers 5 and 11 belong
to the same residue class modulo `3`

. The numbers ```
5
+ 3n
```

, where `n`

is an integer, also belong
to this residue class.

Suppose, you want to solve an equation ```
ax ≡
b(mod m)
```

, where `a`

, `b`

,
and `m`

are integers and `x`

is
an unknown integer. Such equations are called linear congruence equations.
To solve a linear congruence equation, use the `numlib::lincongruence`

function. This
function returns only the solutions * x < m*.
For example, solve the linear congruence equation

```
56x ≡
77(mod 49)
```

:numlib::lincongruence(56, 77, 49)

A linear congruence equation `ax ≡ b(mod m)`

has
at least one solution if and only if the parameters `a`

, `b`

,
and `m`

satisfy the following condition: ```
b
≡ 0(mod gcd(a, m))
```

. If the parameters of a linear
congruence equation do not satisfy this condition, the equation does
not have a solution. In this case, `numlib::lincongruence`

returns `FAIL`

:

numlib::lincongruence(56, 77, 48)

The Chinese remainder theorem states that if the integers `m`

are pairwise coprime, the system of _{i}(i
= 1, ..., n)`n`

linear
congruences `x ≡ a`

has
a solution. The numbers _{i}(mod m_{i})`m`

are pairwise coprime if the greatest common divisor
of any pair of numbers _{i}(i = 1,
..., n)* m_{i}*,

```
i
≠ j
```

) is 1. The solution is unique up to multiples
of the least common multiple (`ilcm`

) of `numlib::ichrem`

function:numlib::ichrem([3, 1, 10], [6, 5, 13])

The Chinese remainder theorem does not state that the system
of linear congruences is solvable only if numbers * m_{1}*,...

`ilcm`

) of numlib::ichrem([5, 7, 9, 6], [10, 11, 12, 13])

If the numbers are not pairwise coprime, a system of linear
congruences does not always have a solution. For unsolvable systems, `numlib::ichrem`

returns `FAIL`

:

numlib::ichrem([5, 1, 9, 6], [10, 15, 12, 13])

To compute modular square roots * x < m* of
the equation

`x`^{2}≡ a(mod
m)

, use the `numlib::msqrts`

function.
Here the integers `a`

and `m`

must
be coprime. For example, solve the congruence equation `x`^{2}≡
13(mod 17)

:numlib::msqrts(13, 17)

If the congruence does not have any solutions, `numlib::msqrts`

returns
an empty set:

numlib::msqrts(10, 17)

If `a`

and `m`

are not coprime, `numlib::msqrts`

errors:

numlib::msqrts(17, 17)

Error: Arguments must be relative prime. [numlib::msqrts]

If `numlib::msqrts`

cannot
solve a congruence, try using the `numlib::mroots`

function. For more information,
see General
Solver for Congruences.

The Legendre symbol determines the solvability of the congruence `x`

, where ^{2}≡
a(mod m)`m`

is a prime. You can
compute the Legendre symbol only if the modulus is a prime number.
The following table demonstrates the dependency between the value
of the Legendre symbol and solvability of the congruence:

If the Legendre number is... | The congruence... |
---|---|

1 | Has one or more solutions |

0 | Cannot be solved by `numlib::msqrts` .
Try `numlib::mroots` . |

-1 | Has no solution |

MuPAD^{®} implements the Legendre symbol as the `numlib::legendre`

function.
If, and only if, the congruence `x`

is solvable, the Legendre symbol is equal to 1:^{2}≡
a(mod m)

numlib::legendre(12, 13)

numlib::msqrts(12, 13)

If, and only if, the congruence `x`

does not have any solutions, the Legendre symbol
is equal to -1:^{2}≡
a(mod m)

numlib::legendre(11, 13)

numlib::msqrts(11, 13)

If `a`

and `m`

are not coprime,
the Legendre symbol is equal to 0. In this case, `numlib::legendre`

function returns 0,
and `numlib::msqrts`

errors:

numlib::legendre(13, 13)

numlib::msqrts(13, 13)

Error: Arguments must be relative prime. [numlib::msqrts]

You can compute the Legendre symbol only if the modulus is a
prime number. If a congruence has a nonprime odd modulus, you can
compute the Jacobi symbol. The Jacobi symbol determines the unsolvable
congruences `x`

.
You cannot compute the Jacobi symbol if the modulus is an even number.
The following table demonstrates the dependency between the value
of the Jacobi symbol and the solvability of the congruence:^{2}≡ a(mod m)

If the Jacobi number is... | The congruence... |
---|---|

1 | Might have solutions |

0 | Cannot be solved by `numlib::msqrts` .
Try `numlib::mroots` . |

-1 | Has no solutions |

MuPAD implements the Jacobi symbol as the `numlib::jacobi`

function.
If the Jacobi symbol is equal to -1, the congruence does not have
a solution:

numlib::jacobi(19, 21)

numlib::msqrts(19, 21)

If Jacobi symbol is equal to 1, the congruence might have solutions:

numlib::jacobi(16, 21)

numlib::msqrts(16, 21)

However, the value 1 of the Jacobi symbol does not guarantee that the congruence has solutions. For example, the following congruence does not have any solutions:

numlib::jacobi(20, 21)

numlib::msqrts(20, 21)

If `a`

and `m`

are not coprime,
the Jacobi symbol is equal to 0. In this case, `numlib::jacobi`

function returns 0, and `numlib::msqrts`

errors:

numlib::jacobi(18, 21)

numlib::msqrts(18, 21)

Error: Arguments must be relative prime. [numlib::msqrts]

Besides solving a linear congruence or computing modular square
roots, MuPAD also enables you to solve congruences of a more
general type of `P(x) ≡ 0(mod m)`

. Here `P(x)`

is
a univariate or multivariate polynomial. To solve such congruences,
use the `numlib::mroots`

function.
For example, solve the congruence `x`

.
First, define the left side of the congruence as a polynomial by using
the ^{3}+
x^{2}+ x + 1 ≡ 0(mod 3)`poly`

function:

p := poly(x^3 + x^2 + x + 1)

Now, use the `numlib::mroots`

function
to solve the congruence:

numlib::mroots(p, 299)

Using the `numlib::mroots`

function,
you also can solve the congruence for a multivariate polynomial. For
a multivariate polynomial `P( x`

, _{1}, ...,
x_{n})`numlib::mroots`

returns a nested list
as a result. Each inner list contains one solution `x`

. For example, find modular
roots of the following multivariate polynomial:_{1},
..., x_{n}

p := poly(x^3*y^2 + x^2*y + x + y + 1): numlib::mroots(p, 11)

Was this topic helpful?