Advanced Math
conjugate and RootOf
eval
Gröbner Bases
product
Series and Limit Computations
Symbolic Integration
Symbolic Summation
The conjugate command now has extended support for RootOf expressions. You can now find the conjugate of the following:
Indexed RootOfs
RootOf expressions with a numerical selector and real coefficients
The following examples return unevaluated in Maple 2015 and earlier.
conjugateRootOfx5+x2−1,index=2
RootOf⁡_Z5+_Z2−1,index=5
conjugateRootOfx2+ 3x+1,−1+I
RootOf⁡_Z2+3⁢_Z+1,−1−I
For indefinite integrals, sums, and products, as well as for differentiations, the eval command now supports an additive change of the variable.
∫fx ⅆxx=a|f(x)x=x+c
∫f⁡x+cⅆx
∑xfxx=a|f(x)x=x+c
∑xf⁡x+c
f'xx=a|f(x)x=x+c
D⁡f⁡x+c
∏xfxx=a|f(x)x=x+c
∏xf⁡x+c
Maple 2016 includes a new C implementation of the F4 algorithm for computing Gröbner bases, replacing FGb. The new code is generally faster and uses multiple threads. The benchmarks below show real time on a quad core Intel Core i5 4590 3.3 GHz computer using a logarithmic scale. The new code supports primes up to 231−1, an increase over FGb's 16-bit primes. To compute over the rationals, Maple uses Chinese remaindering and rational reconstruction.
Maple 2016 includes improved handling of "product over RootOf" cases. The following example returns unevaluated in Maple 2015 and earlier:
∏R = RootOfz2−2x−RootOfz2−R,z
x4−2
A number of improvements were made to series and limit computations in Maple.
The following series and asymptotic functions were added:
Asymptotic expansions of Airy functions at −∞
Series and asymptotic expansions of hypergeometric functions
Series expansions of abs and signum in the real case
Series expansion of GAMMA function at a symbolic pole
Asymptotic expansion of incomplete GAMMA function w.r.t. the parameter
Asymptotic expansion of Hurwitz Zeta function
Series and asymptotic expansions of harmonic numbers
Series expansions of ln and related functions with a logarithmic branch cut depending on a real parameter were improved.
Finally, limit computations of oscillating functions were improved.
New Series Expansions
The following series expansions could not be computed in earlier versions of Maple:
F1 2 functions at 1
serieshypergeom1,1,−13,1−x,x
−89⁢π⁢3x7/3+3227⁢π⁢3x4/3−1681⁢π⁢3x1/3+47−32729⁢π⁢3⁢x2/3+635⁢x−402187⁢π⁢3⁢x5/3+36455⁢x2−646561⁢π⁢3⁢x8/3+811820⁢x3+O⁡x11/3
The case where the lower parameter minus the sum of the upper parameters is an integer is supported.
serieshypergeom12,32,1,1−x,x
2πx+12⁢−ln⁡x−1+4⁢ln⁡2π+12⁢−38⁢ln⁡x−1316+32⁢ln⁡2π⁢x+12⁢−1564⁢ln⁡x−916+1516⁢ln⁡2π⁢x2+12⁢−1751024⁢ln⁡x−525512288+175256⁢ln⁡2π⁢x3+12⁢−220516384⁢ln⁡x−1129132768+22054096⁢ln⁡2π⁢x4+Ox5
The Γ function at a symbolic pole.
seriesΓx,x=n,3 assuming n∷nonposint
1−1n⁢Γ⁡1−nx−n+Ψ⁡1−n−1n⁢Γ⁡1−n+16⁢π2−1n−12⁢Ψ⁡1,1−n⁢Γ⁡1−n+12⁢Ψ⁡1−n2⁢Γ⁡1−n−1n⁢Γ⁡1−n+Ψ⁡1−n2−1nΓ⁡1−n⁢x−n+−−16⁢Ψ⁡2,1−n⁢Γ⁡1−n−12⁢Ψ⁡1,1−n⁢Ψ⁡1−n⁢Γ⁡1−n−16⁢Ψ⁡1−n3⁢Γ⁡1−n−1n⁢Γ⁡1−n−Ψ⁡1−n⁢12⁢Ψ⁡1,1−n⁢Γ⁡1−n+12⁢Ψ⁡1−n2⁢Γ⁡1−n−1n⁢Γ⁡1−n+16⁢3⁢Ψ⁡1−n2+π2−3⁢Ψ⁡1,1−n⁢Ψ⁡1−n−1nΓ⁡1−n⁢x−n2+7360⁢π4−1n−124⁢Ψ⁡3,1−n⁢Γ⁡1−n+16⁢Ψ⁡2,1−n⁢Ψ⁡1−n⁢Γ⁡1−n+18⁢Ψ⁡1,1−n2⁢Γ⁡1−n+14⁢Ψ⁡1,1−n⁢Ψ⁡1−n2⁢Γ⁡1−n+124⁢Ψ⁡1−n4⁢Γ⁡1−n−1n⁢Γ⁡1−n−Ψ⁡1−n⁢−16⁢Ψ⁡2,1−n⁢Γ⁡1−n−12⁢Ψ⁡1,1−n⁢Ψ⁡1−n⁢Γ⁡1−n−16⁢Ψ⁡1−n3⁢Γ⁡1−n−1n⁢Γ⁡1−n−16⁢3⁢Ψ⁡1−n2+π2−3⁢Ψ⁡1,1−n⁢12⁢Ψ⁡1,1−n⁢Γ⁡1−n+12⁢Ψ⁡1−n2⁢Γ⁡1−nΓ⁡1−n⁢−1n+16⁢Ψ⁡1−n3+Ψ⁡1−n⁢π2−3⁢Ψ⁡1,1−n⁢Ψ⁡1−n+Ψ⁡2,1−n⁢Ψ⁡1−n−1nΓ⁡1−n⁢x−n3+Ox−n4
The harmonic function at a negative integer.
seriesharmonic−1+ n,n
−n−1+16⁢π2⁢n−ζ⁡3⁢n2+190⁢π4⁢n3−ζ⁡5⁢n4+1945⁢π6⁢n5+On6
seriesharmonic−2+ n,2,n
−n−2−1+−2+2⁢ζ⁡3⁢n+−3−130⁢π4⁢n2+−4+4⁢ζ⁡5⁢n3+−5−1189⁢π6⁢n4+−6+6⁢ζ⁡7⁢n5+On6
New Asymptotic Expansions
The following asymptotic expansions could not be computed in earlier versions of Maple:
Airy functions at −∞:
seriesAiryAi x,x=−∞
sin⁡23⁢−x3/2+14⁢π⁢−1x1/4π−548⁢cos⁡23⁢−x3/2+14⁢π⁢−1x7/4π−3854608⁢sin⁡23⁢−x3/2+14⁢π⁢−1x13/4π+85085663552⁢cos⁡23⁢−x3/2+14⁢π⁢−1x19/4π+O⁡−1x25/4
seriesAiryBi1,x,x=−∞
sin⁡23⁢−x3/2+14⁢ππ⁢−1x1/4+748⁢cos⁡23⁢−x3/2+14⁢π⁢−1x5/4π+4554608⁢sin⁡23⁢−x3/2+14⁢π⁢−1x11/4π−95095663552⁢cos⁡23⁢−x3/2+14⁢π⁢−1x17/4π+O⁡−1x23/4
Hypergeometric functions w.r.t. the argument:
asympthypergeom1,−12,,x,x
12⁢I⁢π1x−12⁢I⁢π⁢1x+23⁢x+14⁢I⁢π⁢1x3/2−415⁢x2−112⁢I⁢π⁢1x5/2+8105⁢x3+148⁢I⁢π⁢1x7/2−16945⁢x4−1240⁢I⁢π⁢1x9/2+3210395⁢x5+O⁡1x11/2
asympthypergeom−13,2,x,x
−13⁢1x7/3Γ⁡23−2827⁢1x10/3Γ⁡23−980243⁢1x13/3Γ⁡23−1274006561⁢1x16/3Γ⁡23+O⁡1x19/3⁢ⅇx
asympthypergeom13,1,2,x,x
14⁢3⁢Γ⁡23⁢1x13/12π3/2+23192⁢3⁢Γ⁡23⁢1x19/12π3/2+836355296⁢3⁢Γ⁡23⁢1x25/12π3/2+237325884736⁢3⁢Γ⁡23⁢1x31/12π3/2+9307282751528823808⁢3⁢Γ⁡23⁢1x37/12π3/2+12351135110573383542784⁢3⁢Γ⁡23⁢1x43/12π3/2+O⁡1x49/12⁢ⅇ21x
Cases where the two upper parameters of a hypergeom function differ by an integer are handled as well:
asympthypergeom−32,12, 32, x,x
−14⁢I1x3/2+34⁢I1x−I⁢316⁢ln⁡x+316⁢I⁢π+932+38⁢ln⁡2⁢1x+132⁢I⁢1x3/2+3512⁢I⁢1x5/2+1512⁢I⁢1x7/2+O⁡1x9/2
asympthypergeom−52,32,1,1,−x,x
25675⁢π3/2⁢1x5/2+1609⁢π3/2⁢1x3/2+15π3/2⁢1x+58⁢1xπ3/2+53072⁢−12⁢ln⁡x−12⁢γ−72⁢ln⁡2+47⁢1x3/2π3/2+940960⁢−60⁢ln⁡x−60⁢γ−360⁢ln⁡2+283⁢1x5/2π3/2+O⁡1x7/2
Hypergeometric functions w.r.t. a parameter, when the function is actually elementary:
asympthypergeom k, , zk, k
ⅇz+12⁢ⅇz⁢z2k+ⅇz⁢13⁢z3+18⁢z4k2+ⅇz⁢14⁢z4+16⁢z5+148⁢z6k3+ⅇz⁢15⁢z5+1372⁢z6+124⁢z7+1384⁢z8k4+O⁡1k5
converthypergeomk, , zk,elementary
k−zk−k
Hurwitz ζ and incomplete Γ functions w.r.t. the parameter:
asymptZeta0,13,v,v
11v2/3+12⁢1v1/3−154⁢1v4/3+77290⁢1v10/3+O⁡1v16/3
asymptΓv,x,v
2⁢π⁢1v+112⁢2⁢π⁢1v3/2+1288⁢2⁢π⁢1v5/2−13951840⁢2⁢π⁢1v7/2−5712488320⁢2⁢π⁢1v9/2+163879209018880⁢2⁢π⁢1v11/2+O⁡1v13/21vv⁢ⅇv
The harmonic function:
asymptharmonicn,n
ln⁡n+γ+12⁢n−112⁢n2+1120⁢n4+O⁡1n6
asymptharmonicn,23,n
−11n1/3+ζ⁡23−56⁢1n2/3+2554⁢1n5/3−1027⁢1n8/3+239729⁢1n11/3−220729⁢1n14/3+O⁡1n17/3
Two-sided and One-sided Expansions of abs and signum
Maple can now compute two-sided expansions of signum at finite nonzero points.
seriessignuma+x,x,4 assuming x∷real
signum⁡a+−a⁢ℜ⁡aa3+1a⁢x+a⁢−12⁢a2+32⁢ℜ⁡a2a4a−ℜ⁡aa3⁢x2+a⁢32⁢ℜ⁡aa4−52⁢ℜ⁡a3a6a+−12⁢a2+32⁢ℜ⁡a2a4a⁢x3+Ox4
One-sided expansions of abs and signum at 0 and asymptotic expansions can now also be computed.
seriesabssinx,x assuming x>0
x−16⁢x3+1120⁢x5+Ox7
seriessignumsinx,x assuming x>0
1
asymptabsarccotx2−x11,x
1x2+111⁢x3+1121⁢x4+11331⁢x5+O⁡1x6
Expansions of Functions with a Logarithmic Branch Cut Depending on a Real Parameter
For functions ln, arctan, arccot, arctanh, arccoth, Ei, Ci and Chi depending on a parameter, Maple now computes series expansions that are correct for all real values of the parameter (and for all complex values of the series variable sufficiently close to the expansion point).
serieslna+x,x assuming a∷real
ln⁡a+12⁢I⁢signum⁡a⁢csgn⁡I⁢a+x−csgn⁡I⁢a+x⁢π+1a⁢x−12⁢a2⁢x2+13⁢a3⁢x3−14⁢a4⁢x4+15⁢a5⁢x5+Ox6
serieslna x,x assuming a∷real
ln⁡a+ln⁡x+12⁢I⁢csgn⁡I⁢x−signum⁡a⁢csgn⁡I⁢x⁢π
seriesarctanha+x,x,4 assuming a∷real
−I⁢14⁢csgn⁡−I⁢signum⁡a+I⁢a+I⁢x⁢π⁢1+signum⁡a2−1+I⁢ℜ⁡arctanh⁡a+1−a2+1⁢x−aa2−1⁢−a2+1⁢x2−13⁢I⁢−Ia2−1+4⁢I⁢a2a2−12−a2+1⁢x3+Ox4
seriesChia x,x assuming a∷real
γ+ln⁡a+ln⁡x+12⁢I⁢csgn⁡I⁢x−signum⁡a⁢csgn⁡I⁢x⁢π+14⁢a2⁢x2+196⁢a4⁢x4+Ox6
Limits of Oscillating Functions
Limit computations for functions containing oscillating terms were improved. The following limits could not be computed in Maple 2015 or earlier.
limx→∞ⅇⅈ x tanx
undefined
limx→∞−sinhx BesselJ1,x xcoshx−cosx2
−2π..2π
limx→∞−1x−1x2−1x
0
The results for definite integration of rational functions have been improved. In certain cases when the denominator is of degree 4 or higher, the result is now simpler.
Maple 2016
Maple 2015
∫0∞1x4−x+1 ⅆx
−∑_R=RootOf⁡_Z4−_Z+1ln⁡−_R4⁢_R3−1
−∑_R=RootOf⁡229⁢_Z4+18⁢_Z2−8⁢_Z+1_R⁢ln⁡206164⁢_R3+68764⁢_R2+39164⁢_R−2764
∫0∞1x4+6 x+13 ⅆx
153017118852964−318852964⁢∑_R=RootOf⁡_Z4+6⁢_Z+11944⁢_R2−33429⁢_R+24169⁢ln⁡−_R2⁢_R3+3
153017118852964−39426482⁢∑_R=RootOf⁡8684⁢_Z4−194281841538⁢_Z2+201567506232261⁢_Z−1531566190081040464_R⁢ln⁡162056108449387195299530267357790657410141532829⁢_R3+253002615879401450445845355761303849567084999⁢_R2−125011682336520917573411523898216080210433763517237417267771759⁢_R−1530662817658400954319300345845355761303849567084999
In addition, Maple can now compute more definite integrals that could not be computed in Maple 2015 or earlier.
∫0∞π2−arctanx9 ⅆx
−12⁢∑_R=RootOf⁡_Z6−_Z3+1_R3+1⁢ln⁡−_R_R⁢2⁢_R3−1−13⁢π⁢3
∫0∞cosx⋅BesselJ3,x ⅆx
∞
∫−111+1x5+x2+2+1(x5+x2+2)3 ⅆx
25643141911260823328−∑_R=RootOf⁡_Z5+_Z2+2ln⁡−1−_R_R⁢5⁢_R3+2−31260823328⁢∑_R=RootOf⁡_Z5+_Z2+2675729⁢_R3−19964900⁢_R2−5401458⁢_R+75189729⁢ln⁡−1−_R_R⁢5⁢_R3+2+∑_R=RootOf⁡_Z5+_Z2+2ln⁡1−_R_R⁢5⁢_R3+2+31260823328⁢∑_R=RootOf⁡_Z5+_Z2+2675729⁢_R3−19964900⁢_R2−5401458⁢_R+75189729⁢ln⁡1−_R_R⁢5⁢_R3+2
Maple 2016 includes a number of improvements to Maple's symbolic summation engine:
Improved handling of definite parametric sums
New option formal for sum
Support for Jacobi Theta sums
Support for piecewise expressions with more than two branches
Improved divergence testing for infinite sums
Better support for doubly infinite sums
Parametric Sums and Option Formal
Maple 2016 includes several improvements for parametric sums:
The scope of the option parametric was extended so it now works for more types of definite sums.
For infinite sums, Maple is now more careful regarding potentially divergent parametric sums. The behavior can be controlled using assumptions, the _EnvFormal environment variable, or, equivalently, a new option formal to the sum command.
By default, Maple returns a generic answer for certain types of parametric definite hypergeometric sums.
With option parametric, a complete case distinction is now returned for hypergeometric sums with a single parameter that is valid for all integer values of the parameter:
sum−1k binomialm,k k,k=0..m
sumbinomialn,4 k,k=0..n
14⁢2n+14⁢1+In+14⁢1−In
sum−1k binomialm,k k,k=0..m,parametric
{0m≤0−1m=102≤m
sumbinomialn,4 k,k=0..n,parametric
{∑k=0nbinomial⁡n,4⁢kn≤−11n=014⁢2n+14⁢1+In+14⁢1−In1≤n
The behavior for infinite parametric sums of geometric, hypergeometric, polylog, or Zeta type has changed.
Without any assumptions on the parameter, such sums now return unevaluated.
The same sums with appropriate assumptions:
∑k=0∞k2xk
∑k=0∞k2⁢xk
∑k=0∞binomialn+k,k xk
∑k=0∞binomial⁡n+k,k⁢xk
∑k=1∞xkk2
∑k=1∞kn
∑k=0∞k2xk assuming −1<x<1
−x⁢x+1x−13
∑k=0∞k2xk assuming x>1
∑k=0∞binomialn+k,k xk assuming −1<x<1
−−1x−1nx−1
∑k=1∞xkk2 assuming −1<x<1
polylog⁡2,x
∑k=1∞kn assuming Ren <−1
ζ⁡−n
∑k=1∞kn assuming n ≥−1
Alternatively, formal answers can be obtained by either setting the environment variable _EnvFormal≔true, or by specifying the new option formal. (This even works for non-parametric divergent sums.)
For geometric, polylog, and Zeta type sums, option parametric can also be used:
_EnvFormal ≔ true:
∑k=0∞k2 xk
−1+x⁢xx−13
∑k=0∞binomialn+k,k xk
1−x+1n+1
_EnvFormal≔'_EnvFormal':
sumxkk2,k=1..∞
sumxkk2,k=1..∞,formal
sumk3,k=1..∞,formal=ζ⁡−3
1120=1120
sumk2xk,k=0..∞,parametric
{−1+x⁢xx−13x<1∞1≤xundefinedotherwise
sumxkk2,k=1..infinity,parametric
{polylog⁡2,xx=−1Orx=1Orx<1∞1≤xundefinedotherwise
sumkn,k=1..infinity,parametric
{ζ⁡−n1<−ℜ⁡n∞−n≤1undefinedotherwise
Jacobi Theta
Maple now recognizes infinite sums that can be expressed in terms of Jacobi Theta functions.
∑n=0∞rn2 assuming 0<r<1
12⁢JacobiTheta3⁡0,r+12
sumrn2,n=0..∞,parametric
{12⁢JacobiTheta3⁡0,r+12r<1∞1≤rundefinedotherwise
∑n=1∞2−n2−ncos2 n+1⋅t
12⁢21/4⁢JacobiTheta2⁡t,12−cos⁡t
Piecewise Sums
Maple now supports piecewise summands with integer branch points and more than two branches.
∑n=0∞0n≤0an≤12−notherwise
a+12
∑n=0∞n2n=010nn<10xn=10xnotherwiseassuming −1<x<1
9876543210−x11⁢10⁢x−11x−12+10⁢x
Sums Diverging to ±∞
For some non-hypergeometric infinite sums without parameters, Maple now detects when they diverge to ±∞.
∑n=1∞1−1nn
∑n=0∞ cosh2 n+1⋅πcos2 n+1⋅π
−∞
Doubly Infinite Sums
Maple now has improved support for doubly infinite sums, by splitting them into two one-sided infinite sums.
f ≔ i→−1i4 i2−1:
∑i=−∞∞fi
−12⁢π
∑i=−∞−1fi
12−14⁢π
∑i=1∞fi
+f0+
Download Help Document