  
  [1X15 [33X[0;0YNumber Theory[133X[101X
  
  [33X[0;0Y[5XGAP[105X  provides  a  couple  of  elementary number theoretic functions. Most of
  these deal with the group of integers coprime to [22Xm[122X, called the [13Xprime residue
  group[113X.  The  order  of  this  group  is  [22Xϕ(m)[122X  (see [2XPhi[102X  ([14X15.2-2[114X)), and [22Xλ(m)[122X
  (see [2XLambda[102X ([14X15.2-3[114X)) is its exponent. This group is cyclic if and only if [22Xm[122X
  is  2, 4, an odd prime power [22Xp^n[122X, or twice an odd prime power [22X2 p^n[122X. In this
  case  the  generators of the group, i.e., elements of order [22Xϕ(m)[122X, are called
  [13Xprimitive roots[113X (see [2XPrimitiveRootMod[102X ([14X15.3-4[114X)).[133X
  
  [33X[0;0YNote  that  neither  the  arguments  nor  the return values of the functions
  listed below are groups or group elements in the sense of [5XGAP[105X. The arguments
  are simply integers.[133X
  
  
  [1X15.1 [33X[0;0YInfoNumtheor (Info Class)[133X[101X
  
  [1X15.1-1 InfoNumtheor[101X
  
  [33X[1;0Y[29X[2XInfoNumtheor[102X [32X info class[133X
  
  [33X[0;0Y[2XInfoNumtheor[102X  is  the  info  class (see [14X7.4[114X) for the functions in the number
  theory chapter.[133X
  
  
  [1X15.2 [33X[0;0YPrime Residues[133X[101X
  
  [1X15.2-1 PrimeResidues[101X
  
  [33X[1;0Y[29X[2XPrimeResidues[102X( [3Xm[103X ) [32X function[133X
  
  [33X[0;0Y[2XPrimeResidues[102X returns the set of integers from the range [10X[ 0 .. Abs( [3Xm[103X[10X )-1 ][110X
  that are coprime to the integer [3Xm[103X.[133X
  
  [33X[0;0Y[10XAbs([3Xm[103X[10X)[110X must be less than [22X2^28[122X, otherwise the set would probably be too large
  anyhow.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XPrimeResidues( 0 );  PrimeResidues( 1 );  PrimeResidues( 20 );[127X[104X
    [4X[28X[  ][128X[104X
    [4X[28X[ 0 ][128X[104X
    [4X[28X[ 1, 3, 7, 9, 11, 13, 17, 19 ][128X[104X
  [4X[32X[104X
  
  [1X15.2-2 Phi[101X
  
  [33X[1;0Y[29X[2XPhi[102X( [3Xm[103X ) [32X operation[133X
  
  [33X[0;0Y[2XPhi[102X  returns  the  number  [22Xϕ([3Xm[103X)[122X  of positive integers less than the positive
  integer [3Xm[103X that are coprime to [3Xm[103X.[133X
  
  [33X[0;0YSuppose  that  [22Xm = p_1^{e_1} p_2^{e_2} ⋯ p_k^{e_k}[122X. Then [22Xϕ(m)[122X is [22Xp_1^{e_1-1}
  (p_1-1) p_2^{e_2-1} (p_2-1) ⋯ p_k^{e_k-1} (p_k-1)[122X.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XPhi( 12 );[127X[104X
    [4X[28X4[128X[104X
    [4X[25Xgap>[125X [27XPhi( 2^13-1 );  # this proves that 2^(13)-1 is a prime[127X[104X
    [4X[28X8190[128X[104X
    [4X[25Xgap>[125X [27XPhi( 2^15-1 );[127X[104X
    [4X[28X27000[128X[104X
  [4X[32X[104X
  
  [1X15.2-3 Lambda[101X
  
  [33X[1;0Y[29X[2XLambda[102X( [3Xm[103X ) [32X operation[133X
  
  [33X[0;0Y[2XLambda[102X  returns  the exponent [22Xλ([3Xm[103X)[122X of the group of prime residues modulo the
  integer [3Xm[103X.[133X
  
  [33X[0;0Y[22Xλ([3Xm[103X)[122X  is  the  smallest  positive integer [22Xl[122X such that for every [22Xa[122X relatively
  prime  to [3Xm[103X we have [22Xa^l ≡ 1 mod [3Xm[103X[122X. Fermat's theorem asserts [22Xa^{ϕ([3Xm[103X)} ≡ 1 mod
  [3Xm[103X[122X; thus [22Xλ([3Xm[103X)[122X divides [22Xϕ([3Xm[103X)[122X (see [2XPhi[102X ([14X15.2-2[114X)).[133X
  
  [33X[0;0YCarmichael's  theorem  states  that  [22Xλ[122X can be computed as follows: [22Xλ(2) = 1[122X,
  [22Xλ(4)  =  2[122X and [22Xλ(2^e) = 2^{e-2}[122X if [22X3 ≤ e[122X, [22Xλ(p^e) = (p-1) p^{e-1}[122X (i.e. [22Xϕ(m)[122X)
  if [22Xp[122X is an odd prime and [22Xλ(m*n) =[122X[10XLcm[110X[22X( λ(m), λ(n) )[122X if [22Xm, n[122X are coprime.[133X
  
  [33X[0;0YComposites  for  which  [22Xλ(m)[122X  divides [22Xm - 1[122X are called Carmichaels. If [22X6k+1[122X,
  [22X12k+1[122X  and  [22X18k+1[122X  are primes their product is such a number. There are only
  1547 Carmichaels below [22X10^10[122X but 455052511 primes.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XLambda( 10 );[127X[104X
    [4X[28X4[128X[104X
    [4X[25Xgap>[125X [27XLambda( 30 );[127X[104X
    [4X[28X4[128X[104X
    [4X[25Xgap>[125X [27XLambda( 561 );  # 561 is the smallest Carmichael number[127X[104X
    [4X[28X80[128X[104X
  [4X[32X[104X
  
  [1X15.2-4 GeneratorsPrimeResidues[101X
  
  [33X[1;0Y[29X[2XGeneratorsPrimeResidues[102X( [3Xn[103X ) [32X function[133X
  
  [33X[0;0YLet  [3Xn[103X  be a positive integer. [2XGeneratorsPrimeResidues[102X returns a description
  of generators of the group of prime residues modulo [3Xn[103X. The return value is a
  record with components[133X
  
  [8X[10Xprimes[110X[8X: [108X
        [33X[0;6Ya list of the prime factors of [3Xn[103X,[133X
  
  [8X[10Xexponents[110X[8X: [108X
        [33X[0;6Ya list of the exponents of these primes in the factorization of [3Xn[103X, and[133X
  
  [8X[10Xgenerators[110X[8X: [108X
        [33X[0;6Ya  list  describing generators of the group of prime residues; for the
        prime factor [22X2[122X, either a primitive root or a list of two generators is
        stored, for each other prime factor of [3Xn[103X, a primitive root is stored.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XGeneratorsPrimeResidues( 1 );[127X[104X
    [4X[28Xrec( exponents := [  ], generators := [  ], primes := [  ] )[128X[104X
    [4X[25Xgap>[125X [27XGeneratorsPrimeResidues( 4*3 );[127X[104X
    [4X[28Xrec( exponents := [ 2, 1 ], generators := [ 7, 5 ],[128X[104X
    [4X[28X  primes := [ 2, 3 ] )[128X[104X
    [4X[25Xgap>[125X [27XGeneratorsPrimeResidues( 8*9*5 );[127X[104X
    [4X[28Xrec( exponents := [ 3, 2, 1 ],[128X[104X
    [4X[28X  generators := [ [ 271, 181 ], 281, 217 ], primes := [ 2, 3, 5 ] )[128X[104X
  [4X[32X[104X
  
  
  [1X15.3 [33X[0;0YPrimitive Roots and Discrete Logarithms[133X[101X
  
  [1X15.3-1 OrderMod[101X
  
  [33X[1;0Y[29X[2XOrderMod[102X( [3Xn[103X, [3Xm[103X[, [3Xbound[103X] ) [32X function[133X
  
  [33X[0;0Y[2XOrderMod[102X  returns  the  multiplicative  order  of  the  integer [3Xn[103X modulo the
  positive integer [3Xm[103X. If [3Xn[103X and [3Xm[103X are not coprime the order of [3Xn[103X is not defined
  and [2XOrderMod[102X will return [10X0[110X.[133X
  
  [33X[0;0YIf  [3Xn[103X  and  [3Xm[103X are relatively prime the multiplicative order of [3Xn[103X modulo [3Xm[103X is
  the  smallest  positive  integer  [22Xi[122X such that [22X[3Xn[103X^i ≡ 1 mod [3Xm[103X[122X. If the group of
  prime  residues  modulo  [3Xm[103X  is  cyclic then each element of maximal order is
  called a primitive root modulo [3Xm[103X (see [2XIsPrimitiveRootMod[102X ([14X15.3-5[114X)).[133X
  
  [33X[0;0YIf  no a priori known multiple [3Xbound[103X of the desired order is given, [2XOrderMod[102X
  usually  spends  most of its time factoring [3Xm[103X for computing [22Xλ([3Xm[103X)[122X (see [2XLambda[102X
  ([14X15.2-3[114X)) as the default for [3Xbound[103X, and then factoring [3Xbound[103X (see [2XFactorsInt[102X
  ([14X14.4-7[114X)).[133X
  
  [33X[0;0YIf an incorrect [3Xbound[103X is given then the result will be wrong.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XOrderMod( 2, 7 );[127X[104X
    [4X[28X3[128X[104X
    [4X[25Xgap>[125X [27XOrderMod( 3, 7 );  # 3 is a primitive root modulo 7[127X[104X
    [4X[28X6[128X[104X
    [4X[25Xgap>[125X [27Xm:= (5^166-1) / 167;;   # about 10^113[127X[104X
    [4X[25Xgap>[125X [27XOrderMod( 5, m, 166 );  # needs minutes without third argument[127X[104X
    [4X[28X166[128X[104X
  [4X[32X[104X
  
  [1X15.3-2 LogMod[101X
  
  [33X[1;0Y[29X[2XLogMod[102X( [3Xn[103X, [3Xr[103X, [3Xm[103X ) [32X function[133X
  [33X[1;0Y[29X[2XLogModShanks[102X( [3Xn[103X, [3Xr[103X, [3Xm[103X ) [32X function[133X
  
  [33X[0;0Ycomputes  the discrete [3Xr[103X-logarithm of the integer [3Xn[103X modulo the integer [3Xm[103X. It
  returns  a  number  [3Xl[103X  such  that  [22X[3Xr[103X^[3Xl[103X  ≡  [3Xn[103X  mod [3Xm[103X[122X if such a number exists.
  Otherwise [9Xfail[109X is returned.[133X
  
  [33X[0;0Y[2XLogModShanks[102X  uses  the  Baby  Step  -  Giant Step Method of Shanks (see for
  example  [Coh93,  section 5.4.1]) and in general requires more memory than a
  call to [2XLogMod[102X.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27Xl:= LogMod( 2, 5, 7 );  5^l mod 7 = 2;[127X[104X
    [4X[28X4[128X[104X
    [4X[28Xtrue[128X[104X
    [4X[25Xgap>[125X [27XLogMod( 1, 3, 3 );  LogMod( 2, 3, 3 );[127X[104X
    [4X[28X0[128X[104X
    [4X[28Xfail[128X[104X
  [4X[32X[104X
  
  [1X15.3-3 DLog[101X
  
  [33X[1;0Y[29X[2XDLog[102X( [3Xbase[103X, [3Xx[103X[, [3Xm[103X] ) [32X function[133X
  [6XReturns:[106X  [33X[0;10Yan integer[133X
  
  [33X[0;0YThe  argument  [3Xbase[103X  must  be a multiplicative element and [3Xx[103X must lie in the
  cyclic  group  generated  by [3Xbase[103X. The third argument [3Xm[103X must be the order of
  [3Xbase[103X  or  its  factorization.  If [3Xm[103X is not given, it is computed first. This
  function  returns  the  discrete  logarithm,  that is an integer [22Xe[122X such that
  [3Xbase[103X[22X^e =[122X [3Xx[103X.[133X
  
  [33X[0;0YIf [3Xm[103X is prime then Shanks' algorithm is used (which needs [22XO(sqrt[3Xm[103X)[122X space and
  time).  Otherwise  let  [3Xm[103X  [22X=  r  l[122X  and [22Xe = a + b r[122X with [22X0 ≤ a < r[122X. Then [22Xa =[122X
  [10XDLog[110X[22X([3Xbase[103X^l, [3Xx[103X^l, r)[122X and [22Xb =[122X [10XDLog[110X[22X([3Xbase[103X^r, [3Xx[103X/[3Xbase[103X^a, l)[122X.[133X
  
  [33X[0;0YThis function is used for a method of [2XLogFFE[102X ([14X59.2-2[114X).[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27Xq:= 67^12;[127X[104X
    [4X[28X8182718904632857144561[128X[104X
    [4X[25Xgap>[125X [27Xz:= Z(q);;[127X[104X
    [4X[25Xgap>[125X [27XDLog(z, z+1);[127X[104X
    [4X[28X2874413785388345993274[128X[104X
    [4X[25Xgap>[125X [27XDLog(z, z^2+1);[127X[104X
    [4X[28X1667375214152688471247[128X[104X
    [4X[25Xgap>[125X [27XDLog(z, Z(67));[127X[104X
    [4X[28X123980589464134199160[128X[104X
  [4X[32X[104X
  
  [1X15.3-4 PrimitiveRootMod[101X
  
  [33X[1;0Y[29X[2XPrimitiveRootMod[102X( [3Xm[103X[, [3Xstart[103X] ) [32X function[133X
  
  [33X[0;0Y[2XPrimitiveRootMod[102X  returns  the  smallest  primitive root modulo the positive
  integer  [3Xm[103X and [9Xfail[109X if no such primitive root exists. If the optional second
  integer  argument  [3Xstart[103X  is  given  [2XPrimitiveRootMod[102X  returns  the smallest
  primitive root that is strictly larger than [3Xstart[103X.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27X# largest primitive root for a prime less than 2000:[127X[104X
    [4X[25Xgap>[125X [27XPrimitiveRootMod( 409 );[127X[104X
    [4X[28X21[128X[104X
    [4X[25Xgap>[125X [27XPrimitiveRootMod( 541, 2 );[127X[104X
    [4X[28X10[128X[104X
    [4X[25Xgap>[125X [27X# 327 is the largest primitive root mod 337:[127X[104X
    [4X[25Xgap>[125X [27XPrimitiveRootMod( 337, 327 );[127X[104X
    [4X[28Xfail[128X[104X
    [4X[25Xgap>[125X [27X# there exists no primitive root modulo 30:[127X[104X
    [4X[25Xgap>[125X [27XPrimitiveRootMod( 30 );[127X[104X
    [4X[28Xfail[128X[104X
  [4X[32X[104X
  
  [1X15.3-5 IsPrimitiveRootMod[101X
  
  [33X[1;0Y[29X[2XIsPrimitiveRootMod[102X( [3Xr[103X, [3Xm[103X ) [32X function[133X
  
  [33X[0;0Y[2XIsPrimitiveRootMod[102X  returns [9Xtrue[109X if the integer [3Xr[103X is a primitive root modulo
  the  positive  integer [3Xm[103X, and [9Xfalse[109X otherwise. If [3Xr[103X is less than 0 or larger
  than [3Xm[103X it is replaced by its remainder.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XIsPrimitiveRootMod( 2, 541 );[127X[104X
    [4X[28Xtrue[128X[104X
    [4X[25Xgap>[125X [27XIsPrimitiveRootMod( -539, 541 );  # same computation as above;[127X[104X
    [4X[28Xtrue[128X[104X
    [4X[25Xgap>[125X [27XIsPrimitiveRootMod( 4, 541 );[127X[104X
    [4X[28Xfalse[128X[104X
    [4X[25Xgap>[125X [27XForAny( [1..29], r -> IsPrimitiveRootMod( r, 30 ) );[127X[104X
    [4X[28Xfalse[128X[104X
    [4X[25Xgap>[125X [27X# there is no a primitive root modulo 30[127X[104X
  [4X[32X[104X
  
  
  [1X15.4 [33X[0;0YRoots Modulo Integers[133X[101X
  
  [1X15.4-1 Jacobi[101X
  
  [33X[1;0Y[29X[2XJacobi[102X( [3Xn[103X, [3Xm[103X ) [32X function[133X
  
  [33X[0;0Y[2XJacobi[102X  returns  the  value  of  the  [13XKronecker-Jacobi  symbol[113X [22XJ([3Xn[103X,[3Xm[103X)[122X of the
  integer [3Xn[103X modulo the integer [3Xm[103X. It is defined as follows:[133X
  
  [33X[0;0YIf  [22Xn[122X  and  [22Xm[122X  are  not coprime then [22XJ(n,m) = 0[122X. Furthermore, [22XJ(n,1) = 1[122X and
  [22XJ(n,-1)  = -1[122X if [22Xm < 0[122X and [22X+1[122X otherwise. And for odd [22Xn[122X it is [22XJ(n,2) = (-1)^k[122X
  with  [22Xk  =  (n^2-1)/8[122X.  For  odd  primes  [22Xm[122X  which  are  coprime  to  [22Xn[122X  the
  Kronecker-Jacobi   symbol   has  the  same  value  as  the  Legendre  symbol
  (see [2XLegendre[102X ([14X15.4-2[114X)).[133X
  
  [33X[0;0YFor the general case suppose that [22Xm = p_1 ⋅ p_2 ⋯ p_k[122X is a product of [22X-1[122X and
  of primes, not necessarily distinct, and that [22Xn[122X is coprime to [22Xm[122X. Then [22XJ(n,m)
  = J(n,p_1) ⋅ J(n,p_2) ⋯ J(n,p_k)[122X.[133X
  
  [33X[0;0YNote  that the Kronecker-Jacobi symbol coincides with the Jacobi symbol that
  is  defined  for  odd  [22Xm[122X in many number theory books. For odd primes [22Xm[122X and [22Xn[122X
  coprime to [22Xm[122X it coincides with the Legendre symbol.[133X
  
  [33X[0;0Y[2XJacobi[102X  is  very efficient, even for large values of [3Xn[103X and [3Xm[103X, it is about as
  fast as the Euclidean algorithm (see [2XGcd[102X ([14X56.7-1[114X)).[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XJacobi( 11, 35 );  # 9^2 = 11 mod 35[127X[104X
    [4X[28X1[128X[104X
    [4X[25Xgap>[125X [27X# this is -1, thus there is no r such that r^2 = 6 mod 35[127X[104X
    [4X[25Xgap>[125X [27XJacobi( 6, 35 );[127X[104X
    [4X[28X-1[128X[104X
    [4X[25Xgap>[125X [27X# this is 1 even though there is no r with r^2 = 3 mod 35[127X[104X
    [4X[25Xgap>[125X [27XJacobi( 3, 35 );[127X[104X
    [4X[28X1[128X[104X
  [4X[32X[104X
  
  [1X15.4-2 Legendre[101X
  
  [33X[1;0Y[29X[2XLegendre[102X( [3Xn[103X, [3Xm[103X ) [32X function[133X
  
  [33X[0;0Y[2XLegendre[102X  returns  the  value of the [13XLegendre symbol[113X of the integer [3Xn[103X modulo
  the positive integer [3Xm[103X.[133X
  
  [33X[0;0YThe  value  of  the  Legendre symbol [22XL(n/m)[122X is 1 if [22Xn[122X is a [13Xquadratic residue[113X
  modulo  [22Xm[122X, i.e., if there exists an integer [22Xr[122X such that [22Xr^2 ≡ n mod m[122X and [22X-1[122X
  otherwise.[133X
  
  [33X[0;0YIf a root of [3Xn[103X exists it can be found by [2XRootMod[102X ([14X15.4-3[114X).[133X
  
  [33X[0;0YWhile  the  value  of  the  Legendre  symbol usually is only defined for [3Xm[103X a
  prime,  we have extended the definition to include composite moduli too. The
  Jacobi  symbol  (see  [2XJacobi[102X  ([14X15.4-1[114X))  is  another  generalization  of the
  Legendre  symbol  for  composite  moduli  that  is  much cheaper to compute,
  because it does not need the factorization of [3Xm[103X (see [2XFactorsInt[102X ([14X14.4-7[114X)).[133X
  
  [33X[0;0YA  description of the Jacobi symbol, the Legendre symbol, and related topics
  can be found in [Bak84].[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XLegendre( 5, 11 );  # 4^2 = 5 mod 11[127X[104X
    [4X[28X1[128X[104X
    [4X[25Xgap>[125X [27X# this is -1, thus there is no r such that r^2 = 6 mod 11[127X[104X
    [4X[25Xgap>[125X [27XLegendre( 6, 11 );[127X[104X
    [4X[28X-1[128X[104X
    [4X[25Xgap>[125X [27X# this is -1, thus there is no r such that r^2 = 3 mod 35[127X[104X
    [4X[25Xgap>[125X [27XLegendre( 3, 35 );[127X[104X
    [4X[28X-1[128X[104X
  [4X[32X[104X
  
  [1X15.4-3 RootMod[101X
  
  [33X[1;0Y[29X[2XRootMod[102X( [3Xn[103X[, [3Xk[103X], [3Xm[103X ) [32X function[133X
  
  [33X[0;0Y[2XRootMod[102X  computes a [3Xk[103Xth root of the integer [3Xn[103X modulo the positive integer [3Xm[103X,
  i.e.,  a  [22Xr[122X  such that [22Xr^[3Xk[103X ≡ [3Xn[103X mod [3Xm[103X[122X. If no such root exists [2XRootMod[102X returns
  [9Xfail[109X. If only the arguments [3Xn[103X and [3Xm[103X are given, the default value for [3Xk[103X is [22X2[122X.[133X
  
  [33X[0;0YA square root of [3Xn[103X exists only if [10XLegendre([3Xn[103X[10X,[3Xm[103X[10X) = 1[110X (see [2XLegendre[102X ([14X15.4-2[114X)).
  If  [3Xm[103X  has [22Xr[122X different prime factors then there are [22X2^r[122X different roots of [3Xn[103X
  mod  [3Xm[103X.  It  is unspecified which one [2XRootMod[102X returns. You can, however, use
  [2XRootsMod[102X ([14X15.4-4[114X) to compute the full set of roots.[133X
  
  [33X[0;0Y[2XRootMod[102X  is  efficient  even for large values of [3Xm[103X, in fact the most time is
  usually spent factoring [3Xm[103X (see [2XFactorsInt[102X ([14X14.4-7[114X)).[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27X# note 'RootMod' does not return 8 in this case but -8:[127X[104X
    [4X[25Xgap>[125X [27XRootMod( 64, 1009 );[127X[104X
    [4X[28X1001[128X[104X
    [4X[25Xgap>[125X [27XRootMod( 64, 3, 1009 );[127X[104X
    [4X[28X518[128X[104X
    [4X[25Xgap>[125X [27XRootMod( 64, 5, 1009 );[127X[104X
    [4X[28X656[128X[104X
    [4X[25Xgap>[125X [27XList( RootMod( 64, 1009 ) * RootsUnityMod( 1009 ),[127X[104X
    [4X[25X>[125X [27X      x -> x mod 1009 );  # set of all square roots of 64 mod 1009[127X[104X
    [4X[28X[ 1001, 8 ][128X[104X
  [4X[32X[104X
  
  [1X15.4-4 RootsMod[101X
  
  [33X[1;0Y[29X[2XRootsMod[102X( [3Xn[103X[, [3Xk[103X], [3Xm[103X ) [32X function[133X
  
  [33X[0;0Y[2XRootsMod[102X  computes the set of [3Xk[103Xth roots of the integer [3Xn[103X modulo the positive
  integer  [3Xm[103X,  i.e.,  the  list  of all [22Xr[122X such that [22Xr^[3Xk[103X ≡ [3Xn[103X mod [3Xm[103X[122X. If only the
  arguments [3Xn[103X and [3Xm[103X are given, the default value for [3Xk[103X is [22X2[122X.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XRootsMod( 1, 7*31 );  # the same as `RootsUnityMod( 7*31 )'[127X[104X
    [4X[28X[ 1, 92, 125, 216 ][128X[104X
    [4X[25Xgap>[125X [27XRootsMod( 7, 7*31 );[127X[104X
    [4X[28X[ 21, 196 ][128X[104X
    [4X[25Xgap>[125X [27XRootsMod( 5, 7*31 );[127X[104X
    [4X[28X[  ][128X[104X
    [4X[25Xgap>[125X [27XRootsMod( 1, 5, 7*31 );[127X[104X
    [4X[28X[ 1, 8, 64, 78, 190 ][128X[104X
  [4X[32X[104X
  
  [1X15.4-5 RootsUnityMod[101X
  
  [33X[1;0Y[29X[2XRootsUnityMod[102X( [[3Xk[103X, ][3Xm[103X ) [32X function[133X
  
  [33X[0;0Y[2XRootsUnityMod[102X  returns  the  set  of [3Xk[103X-th roots of unity modulo the positive
  integer  [3Xm[103X,  i.e., the list of all solutions [22Xr[122X of [22Xr^[3Xk[103X ≡ [3Xn[103X mod [3Xm[103X[122X. If only the
  argument [3Xm[103X is given, the default value for [3Xk[103X is [22X2[122X.[133X
  
  [33X[0;0YIn  general  there are [22X[3Xk[103X^n[122X such roots if the modulus [3Xm[103X has [22Xn[122X different prime
  factors  [22Xp[122X  such  that  [22Xp ≡ 1 mod [3Xk[103X[122X. If [22X[3Xk[103X^2[122X divides [3Xm[103X then there are [22X[3Xk[103X^{n+1}[122X
  such  roots;  and especially if [22X[3Xk[103X = 2[122X and 8 divides [3Xm[103X there are [22X2^{n+2}[122X such
  roots.[133X
  
  [33X[0;0YIn the current implementation [3Xk[103X must be a prime.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XRootsUnityMod( 7*31 );  RootsUnityMod( 3, 7*31 );[127X[104X
    [4X[28X[ 1, 92, 125, 216 ][128X[104X
    [4X[28X[ 1, 25, 32, 36, 67, 149, 156, 191, 211 ][128X[104X
    [4X[25Xgap>[125X [27XRootsUnityMod( 5, 7*31 );[127X[104X
    [4X[28X[ 1, 8, 64, 78, 190 ][128X[104X
    [4X[25Xgap>[125X [27XList( RootMod( 64, 1009 ) * RootsUnityMod( 1009 ),[127X[104X
    [4X[25X>[125X [27X         x -> x mod 1009 );  # set of all square roots of 64 mod 1009[127X[104X
    [4X[28X[ 1001, 8 ][128X[104X
  [4X[32X[104X
  
  
  [1X15.5 [33X[0;0YMultiplicative Arithmetic Functions[133X[101X
  
  [1X15.5-1 Sigma[101X
  
  [33X[1;0Y[29X[2XSigma[102X( [3Xn[103X ) [32X operation[133X
  
  [33X[0;0Y[2XSigma[102X returns the sum of the positive divisors of the nonzero integer [3Xn[103X.[133X
  
  [33X[0;0Y[2XSigma[102X  is  a  multiplicative  arithmetic  function,  i.e.,  if  [22Xn[122X  and [22Xm[122X are
  relatively prime we have that [22Xσ(n ⋅ m) = σ(n) σ(m)[122X.[133X
  
  [33X[0;0YTogether  with  the  formula  [22Xσ(p^k) = (p^{k+1}-1) / (p-1)[122X this allows us to
  compute [22Xσ([3Xn[103X)[122X.[133X
  
  [33X[0;0YIntegers  [3Xn[103X  for  which [22Xσ([3Xn[103X) = 2 [3Xn[103X[122X are called perfect. Even perfect integers
  are  exactly  of the form [22X2^{[3Xn[103X-1}(2^[3Xn[103X-1)[122X where [22X2^[3Xn[103X-1[122X is prime. Primes of the
  form  [22X2^[3Xn[103X-1[122X  are  called  [13XMersenne  primes[113X,  and 42 among the known Mersenne
  primes  are  obtained  for [3Xn[103X [22X=[122X 2, 3, 5, 7, 13, 17, 19, 31, 61, 89, 107, 127,
  521,  607,  1279,  2203,  2281,  3217, 4253, 4423, 9689, 9941, 11213, 19937,
  21701, 23209, 44497, 86243, 110503, 132049, 216091, 756839, 859433, 1257787,
  1398269,   2976221,  3021377,  6972593,  13466917,  20996011,  24036583  and
  25964951.  Please  find more up to date information about Mersenne primes at
  [7Xhttps://www.mersenne.org[107X.  It  is  not  known  whether  odd perfect integers
  exist,  however [BC89]  show  that  any  such integer must have at least 300
  decimal digits.[133X
  
  [33X[0;0Y[2XSigma[102X usually spends most of its time factoring [3Xn[103X (see [2XFactorsInt[102X ([14X14.4-7[114X)).[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XSigma( 1 );[127X[104X
    [4X[28X1[128X[104X
    [4X[25Xgap>[125X [27XSigma( 1009 );  # 1009 is a prime[127X[104X
    [4X[28X1010[128X[104X
    [4X[25Xgap>[125X [27XSigma( 8128 ) = 2*8128;  # 8128 is a perfect number[127X[104X
    [4X[28Xtrue[128X[104X
  [4X[32X[104X
  
  [1X15.5-2 Tau[101X
  
  [33X[1;0Y[29X[2XTau[102X( [3Xn[103X ) [32X operation[133X
  
  [33X[0;0Y[2XTau[102X returns the number of the positive divisors of the nonzero integer [3Xn[103X.[133X
  
  [33X[0;0Y[2XTau[102X  is  a multiplicative arithmetic function, i.e., if [22Xn[122X and [22Xm[122X are relative
  prime  we  have [22Xτ(n ⋅ m) = τ(n) τ(m)[122X. Together with the formula [22Xτ(p^k) = k+1[122X
  this allows us to compute [22Xτ([3Xn[103X)[122X.[133X
  
  [33X[0;0Y[2XTau[102X usually spends most of its time factoring [3Xn[103X (see [2XFactorsInt[102X ([14X14.4-7[114X)).[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XTau( 1 );[127X[104X
    [4X[28X1[128X[104X
    [4X[25Xgap>[125X [27XTau( 1013 );  # thus 1013 is a prime[127X[104X
    [4X[28X2[128X[104X
    [4X[25Xgap>[125X [27XTau( 8128 );[127X[104X
    [4X[28X14[128X[104X
    [4X[25Xgap>[125X [27X# result is odd if and only if argument is a perfect square:[127X[104X
    [4X[25Xgap>[125X [27XTau( 36 );[127X[104X
    [4X[28X9[128X[104X
  [4X[32X[104X
  
  [1X15.5-3 MoebiusMu[101X
  
  [33X[1;0Y[29X[2XMoebiusMu[102X( [3Xn[103X ) [32X function[133X
  
  [33X[0;0Y[2XMoebiusMu[102X  computes  the value of Moebius inversion function for the nonzero
  integer  [3Xn[103X. This is 0 for integers which are not squarefree, i.e., which are
  divided  by a square [22Xr^2[122X. Otherwise it is 1 if [3Xn[103X has a even number and [22X-1[122X if
  [3Xn[103X has an odd number of prime factors.[133X
  
  [33X[0;0YThe importance of [22Xμ[122X stems from the so called inversion formula. Suppose [22Xf[122X is
  a  multiplicative  arithmetic  function defined on the positive integers and
  let  [22Xg(n)  = ∑_{d ∣ n} f(d)[122X. Then [22Xf(n) = ∑_{d ∣ n} μ(d) g(n/d)[122X. As a special
  case  we  have  [22Xϕ(n)  = ∑_{d ∣ n} μ(d) n/d[122X since [22Xn = ∑_{d ∣ n} ϕ(d)[122X (see [2XPhi[102X
  ([14X15.2-2[114X)).[133X
  
  [33X[0;0Y[2XMoebiusMu[102X  usually  spends  all  of  its  time  factoring  [3Xn[103X (see [2XFactorsInt[102X
  ([14X14.4-7[114X)).[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XMoebiusMu( 60 );  MoebiusMu( 61 );  MoebiusMu( 62 );[127X[104X
    [4X[28X0[128X[104X
    [4X[28X-1[128X[104X
    [4X[28X1[128X[104X
  [4X[32X[104X
  
  
  [1X15.6 [33X[0;0YContinued Fractions[133X[101X
  
  [1X15.6-1 ContinuedFractionExpansionOfRoot[101X
  
  [33X[1;0Y[29X[2XContinuedFractionExpansionOfRoot[102X( [3Xf[103X, [3Xn[103X ) [32X function[133X
  
  [33X[0;0YThe  first  [3Xn[103X terms of the continued fraction expansion of the only positive
  real  root  of  the  polynomial  [3Xf[103X  with  integer  coefficients. The leading
  coefficient  of [3Xf[103X must be positive and the value of [3Xf[103X at 0 must be negative.
  If  the  degree of [3Xf[103X is 2 and [3Xn[103X = 0, the function computes one period of the
  continued fraction expansion of the root in question. Anything may happen if
  [3Xf[103X has three or more positive real roots.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27Xx := Indeterminate(Integers);;[127X[104X
    [4X[25Xgap>[125X [27XContinuedFractionExpansionOfRoot(x^2-7,20);[127X[104X
    [4X[28X[ 2, 1, 1, 1, 4, 1, 1, 1, 4, 1, 1, 1, 4, 1, 1, 1, 4, 1, 1, 1 ][128X[104X
    [4X[25Xgap>[125X [27XContinuedFractionExpansionOfRoot(x^2-7,0);[127X[104X
    [4X[28X[ 2, 1, 1, 1, 4 ][128X[104X
    [4X[25Xgap>[125X [27XContinuedFractionExpansionOfRoot(x^3-2,20);[127X[104X
    [4X[28X[ 1, 3, 1, 5, 1, 1, 4, 1, 1, 8, 1, 14, 1, 10, 2, 1, 4, 12, 2, 3 ][128X[104X
    [4X[25Xgap>[125X [27XContinuedFractionExpansionOfRoot(x^5-x-1,50);[127X[104X
    [4X[28X[ 1, 5, 1, 42, 1, 3, 24, 2, 2, 1, 16, 1, 11, 1, 1, 2, 31, 1, 12, 5,[128X[104X
    [4X[28X  1, 7, 11, 1, 4, 1, 4, 2, 2, 3, 4, 2, 1, 1, 11, 1, 41, 12, 1, 8, 1,[128X[104X
    [4X[28X  1, 1, 1, 1, 9, 2, 1, 5, 4 ][128X[104X
  [4X[32X[104X
  
  [1X15.6-2 ContinuedFractionApproximationOfRoot[101X
  
  [33X[1;0Y[29X[2XContinuedFractionApproximationOfRoot[102X( [3Xf[103X, [3Xn[103X ) [32X function[133X
  
  [33X[0;0YThe  [3Xn[103Xth  continued fraction approximation of the only positive real root of
  the  polynomial  [3Xf[103X  with  integer coefficients. The leading coefficient of [3Xf[103X
  must  be  positive  and  the  value of [3Xf[103X at 0 must be negative. Anything may
  happen if [3Xf[103X has three or more positive real roots.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XContinuedFractionApproximationOfRoot(x^2-2,10);[127X[104X
    [4X[28X3363/2378[128X[104X
    [4X[25Xgap>[125X [27X3363^2-2*2378^2;[127X[104X
    [4X[28X1[128X[104X
    [4X[25Xgap>[125X [27Xz := ContinuedFractionApproximationOfRoot(x^5-x-1,20);[127X[104X
    [4X[28X499898783527/428250732317[128X[104X
    [4X[25Xgap>[125X [27Xz^5-z-1;[127X[104X
    [4X[28X486192462527432755459620441970617283/[128X[104X
    [4X[28X14404247382319842421697357558805709031116987826242631261357[128X[104X
  [4X[32X[104X
  
  
  [1X15.7 [33X[0;0YMiscellaneous[133X[101X
  
  [1X15.7-1 PValuation[101X
  
  [33X[1;0Y[29X[2XPValuation[102X( [3Xn[103X, [3Xp[103X ) [32X function[133X
  
  [33X[0;0YFor  an  integer [3Xn[103X and a prime [3Xp[103X this function returns the [3Xp[103X-valuation of [3Xn[103X,
  that  is the exponent [22Xe[122X such that [22Xp^e[122X is the largest power of [3Xp[103X that divides
  [3Xn[103X. The valuation of zero is infinity.[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XPValuation(100,2);[127X[104X
    [4X[28X2[128X[104X
    [4X[25Xgap>[125X [27XPValuation(100,3);[127X[104X
    [4X[28X0[128X[104X
  [4X[32X[104X
  
  [1X15.7-2 TwoSquares[101X
  
  [33X[1;0Y[29X[2XTwoSquares[102X( [3Xn[103X ) [32X function[133X
  
  [33X[0;0Y[2XTwoSquares[102X  returns  a  list  of two integers [22Xx ≤ y[122X such that the sum of the
  squares  of  [22Xx[122X  and [22Xy[122X is equal to the nonnegative integer [3Xn[103X, i.e., [22Xn = x^2 +
  y^2[122X.   If  no  such  representation  exists  [2XTwoSquares[102X  will  return  [9Xfail[109X.
  [2XTwoSquares[102X  will  return a representation for which the gcd of [22Xx[122X and [22Xy[122X is as
  small  as  possible.  It  is  not  specified which representation [2XTwoSquares[102X
  returns if there is more than one.[133X
  
  [33X[0;0YLet  [22Xa[122X  be  the  product  of  all  maximal powers of primes of the form [22X4k+3[122X
  dividing [3Xn[103X. A representation of [3Xn[103X as a sum of two squares exists if and only
  if  [22Xa[122X is a perfect square. Let [22Xb[122X be the maximal power of [22X2[122X dividing [3Xn[103X or its
  half,  whichever is a perfect square. Then the minimal possible gcd of [22Xx[122X and
  [22Xy[122X  is  the  square  root  [22Xc[122X  of  [22Xa  ⋅  b[122X.  The  number  of different minimal
  representation  with  [22Xx  ≤  y[122X is [22X2^{l-1}[122X, where [22Xl[122X is the number of different
  prime factors of the form [22X4k+1[122X of [3Xn[103X.[133X
  
  [33X[0;0YThe  algorithm  first  finds a square root [22Xr[122X of [22X-1[122X modulo [22X[3Xn[103X / (a ⋅ b)[122X, which
  must  exist,  and  applies  the  Euclidean  algorithm  to [22Xr[122X and [3Xn[103X. The first
  residues in the sequence that are smaller than [22Xsqrt{[3Xn[103X/(a ⋅ b)}[122X times [22Xc[122X are a
  possible pair [22Xx[122X and [22Xy[122X.[133X
  
  [33X[0;0YBetter  descriptions  of  the  algorithm  and related topics can be found in
  [Wag90] and [Zag90].[133X
  
  [4X[32X  Example  [32X[104X
    [4X[25Xgap>[125X [27XTwoSquares( 5 );[127X[104X
    [4X[28X[ 1, 2 ][128X[104X
    [4X[25Xgap>[125X [27XTwoSquares( 11 );  # there is no representation[127X[104X
    [4X[28Xfail[128X[104X
    [4X[25Xgap>[125X [27XTwoSquares( 16 );[127X[104X
    [4X[28X[ 0, 4 ][128X[104X
    [4X[25Xgap>[125X [27X# 3 is the minimal possible gcd because 9 divides 45:[127X[104X
    [4X[25Xgap>[125X [27XTwoSquares( 45 );[127X[104X
    [4X[28X[ 3, 6 ][128X[104X
    [4X[25Xgap>[125X [27X# it is not [5,10] because their gcd is not minimal:[127X[104X
    [4X[25Xgap>[125X [27XTwoSquares( 125 );[127X[104X
    [4X[28X[ 2, 11 ][128X[104X
    [4X[25Xgap>[125X [27X# [10,11] would be the other possible representation:[127X[104X
    [4X[25Xgap>[125X [27XTwoSquares( 13*17 );[127X[104X
    [4X[28X[ 5, 14 ][128X[104X
    [4X[25Xgap>[125X [27XTwoSquares( 848654483879497562821 );  # argument is prime[127X[104X
    [4X[28X[ 6305894639, 28440994650 ][128X[104X
  [4X[32X[104X
  
