@hackage hops0.2.0

Handy Operations on Power Series

HOPS HOPS Build Status

Handy Operations on Power Series.

Install

The easiest way to get started is to download a prebuilt binary. Such binaries can be found on the releases page. The binaries are statically linked and should work on any Linux system.

Alternative ways of installing hops include using the nix package manager:

$ nix-env -f "<nixpkgs>" -iA haskellPackages.hops

Or using cabal:

$ cabal update && cabal install hops

Introduction

To get a feeling for the HOPS language and using its interpreter (hops) let us look at a few examples.

Fibonacci numbers

The generating function, f, for the Fibonacci numbers satisfies f=1+(x+x2)f, and we can get the coefficient of f directly from that equation:

$ hops 'f=1+(x+x^2)*f'
f=1+(x+x^2)*f => {1,1,2,3,5,8,13,21,34,55,89,144,233,377,610}

Alternatively, we could first solve for f in f=1+(x+x2)f and let hops expand that expression:

$ hops 'f=1/(1-x-x^2)'
f=1/(1-x-x^2) => {1,1,2,3,5,8,13,21,34,55,89,144,233,377,610}

Catalan numbers

It could hardly be easier:

$ hops 'C=1+x*C^2'
C=1+x*C^2 => {1,1,2,5,14,42,132,429,1430,4862,16796,58786,208012,742900,2674440}

Bell numbers

The exponential generating function for the Bell numbers is eex-1 and we can give that expression to hops:

$ hops --prec=10 'exp(exp(x)-1)'
exp(exp(x)-1) => {1,1,1,5/6,5/8,13/30,203/720,877/5040,23/224,1007/17280}

To get the Bell numbers we, however, also need to multiply the coefficient of xn in that series by n!; this is what the laplace transform does:

$ hops --prec=10 'f=exp(exp(x)-1);laplace(f)'
f=exp(exp(x)-1);laplace(f) => {1,1,2,5,15,52,203,877,4140,21147}

Euler numbers

Power series defined by trigonometric functions are fine too:

$ hops --prec=12 'f=sec(x)+tan(x);laplace(f)'
f=sec(x)+tan(x);laplace(f) => {1,1,1,2,5,16,61,272,1385,7936,50521,353792}

Number of ballots (ordered set partitions)

This sequence is most simply defined by its exponential generating function y=1/(2-ex):

$ hops --prec 10 'y=1/(2-exp(x));laplace(y)'
y=1/(2-exp(x));laplace(y) => {1,1,3,13,75,541,4683,47293,545835,7087261}

Alternatively, one can exploit that y'=2y2-y:

$ hops --prec 10 'y=1+integral(2*y^2-y);laplace(y)'
y=1+integral(2*y^2-y);laplace(y) => {1,1,3,13,75,541,4683,47293,545835,7087261}

Number of interval orders

Let A be the exponential generating function for the number of labeled interval orders (A079144). Zagier [Vassiliev invariants and a strange identity related to the Dedekind eta-function, Topology 40 (2001)] showed that A(24*x)=exp(x)*T(x) where T is the exponential generating function for the seqence of Glaisher's T numbers (A002439). Moreover, the exponential generating function for the aerated seqence of Glaisher's T numbers is sin(2*x)/(2*cos(3*x)). Putting this together we have an HOPS-expression for the number of labeled interval orders:

$ hops 'T=laplacei(BISECT1(laplace(sin(2*x)/(2*cos(3*x)))));laplace((exp(x)*T)@(x/24))'
T=laplacei(BISECT1(laplace(sin(2*x)/(2*cos(3*x)))));laplace((exp(x)*T)@(x/24)) => {1,1,3,19,207,3451,81663}

Simple sequence notation

We have seen how to define a few different sequences using generating functions and functional equations. HOPS also supports a more naive way of specifying sequences. Here's a simple finite sequence:

$ hops '{1,2,3}'
{1,2,3} => {1,2,3}

We can also use ellipses to build infinite sequences:

$ hops '{1,2,...}'
{1,2,...} => {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}

What happened in the background here is that hops fitted the first degree polynomial p(n)=1+n to the values p(0)=1 and p(1)=2. We could alternatively have given this formula explicitly:

$ hops '{1+n}'
{1+n} => {1,2,3,4,5,6,7,8,9,10,11,12,13,14,15}

We are not limited to first degree polynomials either:

$ hops '{0,1,8,27,...}'
{0,1,8,27,...} => {0,1,8,27,64,125,216,343,512,729,1000,1331,1728,2197,2744}

$ hops '{n^3}'
{n^3} => {0,1,8,27,64,125,216,343,512,729,1000,1331,1728,2197,2744}

The number of integer compositions of n is 1 if n=0 and 2n-1 if n>0; see A011782. Here's how we might specify that formula:

$ hops '{1,2^(n-1)}'
{1,2^(n-1)} => {1,1,2,4,8,16,32,64,128,256,512,1024,2048,4096,8192}

Factorials are fine too. Here's the order of the alternating group (A001710):

$ hops --prec=12 '{1,1,n!/2}'
{1,1,n!/2} => {1,1,1,3,12,60,360,2520,20160,181440,1814400,19958400}

Composing programs

Using the special variable stdin we can compose programs:

$ hops 'f=1+(x+x^2)*f' | hops 'stdin/(1-x)'
f=1+(x+x^2)*f;f/(1-x) => {1,2,4,7,12,20,33,54,88,143,232,376,609,986,1596}

As a side note, one can show that HOPS programs form a monoid under this type of composition.

Be aware that hops may have to rename variables when composing programs:

$ hops --prec=10 'f=1+(x+x^2)*f' | hops 'f=1/(1-2*x);f/(1-x*stdin)'
f=1+(x+x^2)*f;g=1+2*x*g;g/(1-x*f) => {1,3,8,21,54,137,344,857,2122,5229,12836}

OEIS A-numbers

OEIS A-numbers can be used directly in HOPS programs and they are interpreted as ordinary generating functions. E.g. this is the difference between the Catalan numbers (A000108) and the Motzkin numbers (A001006):

$ hops 'A000108-A001006'
A000108-A001006 => {0,0,0,1,5,21,81,302,1107,4027,14608,52988,192501,701065,2560806}

Misc transformations

HOPS knows about many of the transformations used by OEIS https://oeis.org/transforms.html.

As an example, the sequence A067145 claims to shift left under reversion:

S A067145 1,1,-1,3,-13,69,-419,2809,-20353,157199,-1281993,10963825,-97828031,
N A067145 Shifts left under reversion.

Let's test that claim:

$ hops 'REVERT(A067145)-LEFT(A067145)'
REVERT(A067145)-LEFT(A067145) => {0,0,0,0,0,0,0,0,0,0,0,0,0,0,0}

HOPS scripts

Sometimes it is useful be able to apply many transformations to the same input. One way to achieve that is to write a script with the transformations we are interested in. E.g. if we create a file transforms.hops containing

BINOMIAL(stdin)
EULER(stdin)
REVEGF(stdin)
STIRLING(stdin)

then we can apply all of these transforms to 1/(1-x) as follows:

$ hops '1/(1-x)' | hops --prec=9 -f transforms.hops
f=1/(1-x);BINOMIAL(f) => {1,2,4,8,16,32,64,128,256}
f=1/(1-x);EULER(f) => {1,2,3,5,7,11,15,22,30}
f=1/(1-x);REVEGF(f) => {1,-2,9,-64,625,-7776,117649,-2097152,43046721}
f=1/(1-x);STIRLING(f) => {1,2,5,15,52,203,877,4140,21147}

N.B: As in this example, the preferred file extension for HOPS program files is .hops.

The HOPS language

Each line of a HOPS script is an independent program and each line/program consists of a semicolon separated list of functional equations and generating functions. We shall now describe operations, functions and transformations that can be used when building such programs.

Binary operations

Operation Meaning
f + g sum of f and g
f - g difference of f and g
f ^ g f to the power g
f @ g f composed with g (can also be written f(g) when f is a name)
f * g product of f and g
f / g quotient of f and g
f .* g coefficient-wise/Hadamard product of f and g
f ./ g coefficient-wise quotient of f and g

Derivative and integral

Operation Meaning
D(f) derivative of f
integral(f) integral of f

Functions

Function Meaning
sqrt(f) f^(1/2)
abs(f) coefficient-wise absolute value
log(f) logarithmic function
exp(f) exponential function
sin(f) sine function
cos(f) cosine function
tan(f) tangent function
sec(f) 1/cos(f)
arcsin(f) arcsine function
arccos(f) arccosine function
arctan(f) arctangent function
sinh(f) hyperbolic sine function
cosh(f) hyperbolic cosine function
tanh(f) hyperbolic tangent function
arsinh(f) area hyperbolic sine function
arcosh(f) area hyperbolic cosine function
artanh(f) area hyperbolic tangent function
laplace(f) f .* {n!}
laplacei(f) f ./ {n!}
revert(f) the compositional inverse of f

Transforms

Transform Meaning
AERATE1(f) f(x^2)
AERATE2(f) f(x^3)
BARRY1(f) 1/(1-x-x^2*f)
BARRY2(f) 1/(1+x+x^2*f)
BINOMIAL(f) g=exp(x)*laplacei(f);laplace(g)
BINOMIALi(f) g=exp(-x)*laplacei(f);laplace(g)
BIN1(f) g={(-1)^n/n!}*((laplacei(x*f))@(-x));LEFT(laplace(-g))
BISECT0(f) if f={a0,a1,a2,a3,a4,...} then BISECT0(f)={a0,a2,a4,...}
BISECT1(f) if f={a0,a1,a2,a3,a4,...} then BISECT1(f)={a1,a3,a5,...}
BOUS2(f) See [1]
BOUS2i(f) See [1]
BOUS(f) See [1]
CONV(f) f^2
CONVi(f) sqrt(f)
DIFF(f) LEFT(f)-f
EULER(f) Euler transform
EULERi(f) inverse Euler transform
EXPCONV(f) g=laplacei(f);laplace(g*g)
EXP(f) g={1/n!}@(laplacei(x*f));laplace(g-1)/x
HANKEL(f) Hankel transform
lHANKEL(f) g=f.*f-LEFT(f).*RIGHT(f);LEFT(g)
INVERT(f) LEFT(1/(1-x*f))
INVERTi(f) LEFT(-1/(1+x*f))
LAH(f) g=(laplacei(f))@(x/(1-x));laplace(g)
LAHi(f) g=(laplacei(f))@(x/(1+x));laplace(g)
LEFT(f) if f={a0,a1,a2,a3,a4,...} then LEFT(f)={a1,a2,a3,...}
LOG(f) g=log(1+laplacei(x*f));LEFT(laplace(g))
M2(f) 2*f-f(0)
M2i(f) (f + f(0))/2
MOBIUS(f) See [1]
MOBIUSi(f) See [1]
NEGATE(f) (1-x/(1-x)).*f
PARTITION(f) See [1]
POINT(f) laplace(x*D(laplacei(f)))
PRODS(f) if f = {a0,a1,a2,...} then PRODS(f)={a0,a0*a1,a0*a1*a2,...}
PSUM(f) f/(1-x)
PSUMSIGN(f) f/(1+x)
REVERT(f) LEFT(revert(x*f))
REVEGF(f) LEFT(laplace(revert((x*f)./(1+x*laplace(1/(1-x))))))
RIGHT(f) 1+x*f
STIRLING(f) g=laplacei(x*f);laplace(g@({0,1/n!}))/x
STIRLINGi(f) g=laplacei(x*f);laplace(g@({0,(-1)^(n+1)/n!}))/x
T019(f) if f={a[n]} then {a[n+2]-2*a[n+1]+a[n]}
TRISECT0(f) if f={a0,a1,a2,a3,a4,...} then TRISECT0(f)={a0,a3,a6,...}
TRISECT1(f) if f={a0,a1,a2,a3,a4,...} then TRISECT0(f)={a1,a4,a7,...}
TRISECT2(f) if f={a0,a1,a2,a3,a4,...} then TRISECT0(f)={a2,a5,a8,...}
WEIGHT(f) if f={a0,a1,a2,...} then WEIGHT(f)=(1+x^n)^a0*(1+x^n)^a1*...

[1] https://oeis.org/transforms.txt

A grammar for HOPS scripts

A HOPS script is a list of independent programs (prg) - one program per line:

hops = prg { "\n" prg }

A program is a list of semicolon separated commands (cmd):

prg = cmd { ";" cmd }

A command is a generating function expression (expr0) or an assignment:

cmd = expr0 | name "=" expr0

We use the precedence climbing method to define generating function expressions:

expr0 = expr0 ("+" | "-") expr0 | expr1

expr1 = expr1 ("*" | "/" | ".*" | "./") expr1 | expr2

expr2 = ("-" | "+") expr2 | expr3 "!" | expr3 "^" expr3 | expr3 "@" expr3 | expr3

expr3 = "x" | anum | tag | name | lit | "{" { terms } "}" | name "(" expr3 ")" | expr0

lit = int

int = digit { digit }

digit = "0" | "1" | ... | "9"

alpha = "A" | "B" | ... | "Z" | "a" | "b" | ... | "z"

alphanum = alpha | digit

name = alphanum { alphanum | "_" }

terms = cexpr0 { "," expr0 } ("..." | cexpr0 | fun)

fun = the same as cexpr0 except lit = linear

linear = int | int "*n"

cexpr0 = cexpr0 ("+" | "-") cexpr0 | cexpr1

cexpr1 = cexpr1 ("*" | "/") cexpr1 | cexpr2

cexpr2 = ("+" | "-") cexpr2 | cexpr3 "!" | cexpr3 "^" cexpr3 | cexpr3

cexpr3 = lit | cexpr0

The man page

The hops command has additional functionality such as the ability to assign tags to sequences:

$ printf "1,1,2,5,17,33\n1,1,2,5,19,34\n" | hops --tag 1
TAG000001 => {1,1,2,5,17,33}
TAG000002 => {1,1,2,5,19,34}

For further information regarding command line options to hops see the man page.

Issues

Have you found a bug? Want to contribute to hops? Please open an issue at https://github.com/akc/hops/issues.

How to cite

@misc{hops,
  author = "Anders Claesson",
  title  = "HOPS: Handy Operations on Power Series",
  year   =  2015,
  howpublished = "\url{http://akc.is/hops}"
}

License

BSD-3: see the LICENSE file.