Integer compositions

A composition \(c\) of a nonnegative integer \(n\) is a list of positive integers (the parts of the composition) with total sum \(n\).

This module provides tools for manipulating compositions and enumerated sets of compositions.

EXAMPLES:

sage: Composition([5, 3, 1, 3])
[5, 3, 1, 3]
sage: list(Compositions(4))
[[1, 1, 1, 1], [1, 1, 2], [1, 2, 1], [1, 3], [2, 1, 1], [2, 2], [3, 1], [4]]

>>> from sage.all import *
>>> Composition([Integer(5), Integer(3), Integer(1), Integer(3)])
[5, 3, 1, 3]
>>> list(Compositions(Integer(4)))
[[1, 1, 1, 1], [1, 1, 2], [1, 2, 1], [1, 3], [2, 1, 1], [2, 2], [3, 1], [4]]

AUTHORS:

  • Mike Hansen, Nicolas M. Thiéry

  • MuPAD-Combinat developers (algorithms and design inspiration)

  • Travis Scrimshaw (2013-02-03): Removed CombinatorialClass

class sage.combinat.composition.Composition(parent, lst)[source]

Bases: CombinatorialElement

Integer compositions.

A composition of a nonnegative integer \(n\) is a list \((i_1, \ldots, i_k)\) of positive integers with total sum \(n\).

EXAMPLES:

The simplest way to create a composition is by specifying its entries as a list, tuple (or other iterable):

sage: Composition([3,1,2])
[3, 1, 2]
sage: Composition((3,1,2))
[3, 1, 2]
sage: Composition(i for i in range(2,5))
[2, 3, 4]

>>> from sage.all import *
>>> Composition([Integer(3),Integer(1),Integer(2)])
[3, 1, 2]
>>> Composition((Integer(3),Integer(1),Integer(2)))
[3, 1, 2]
>>> Composition(i for i in range(Integer(2),Integer(5)))
[2, 3, 4]

You can also create a composition from its code. The code of a composition \((i_1, i_2, \ldots, i_k)\) of \(n\) is a list of length \(n\) that consists of a \(1\) followed by \(i_1-1\) zeros, then a \(1\) followed by \(i_2-1\) zeros, and so on.

sage: Composition([4,1,2,3,5]).to_code()
[1, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
sage: Composition(code=_)
[4, 1, 2, 3, 5]
sage: Composition([3,1,2,3,5]).to_code()
[1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
sage: Composition(code=_)
[3, 1, 2, 3, 5]

>>> from sage.all import *
>>> Composition([Integer(4),Integer(1),Integer(2),Integer(3),Integer(5)]).to_code()
[1, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
>>> Composition(code=_)
[4, 1, 2, 3, 5]
>>> Composition([Integer(3),Integer(1),Integer(2),Integer(3),Integer(5)]).to_code()
[1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
>>> Composition(code=_)
[3, 1, 2, 3, 5]

You can also create the composition of \(n\) corresponding to a subset of \(\{1, 2, \ldots, n-1\}\) under the bijection that maps the composition \((i_1, i_2, \ldots, i_k)\) of \(n\) to the subset \(\{i_1, i_1 + i_2, i_1 + i_2 + i_3, \ldots, i_1 + \cdots + i_{k-1}\}\) (see to_subset()):

sage: Composition(from_subset=({1, 2, 4}, 5))
[1, 1, 2, 1]
sage: Composition([1, 1, 2, 1]).to_subset()
{1, 2, 4}

>>> from sage.all import *
>>> Composition(from_subset=({Integer(1), Integer(2), Integer(4)}, Integer(5)))
[1, 1, 2, 1]
>>> Composition([Integer(1), Integer(1), Integer(2), Integer(1)]).to_subset()
{1, 2, 4}

The following notation equivalently specifies the composition from the set \(\{i_1 - 1, i_1 + i_2 - 1, i_1 + i_2 + i_3 - 1, \dots, i_1 + \cdots + i_{k-1} - 1, n-1\}\) or \(\{i_1 - 1, i_1 + i_2 - 1, i_1 + i_2 + i_3 - 1, \dots, i_1 + \cdots + i_{k-1} - 1\}\) and \(n\). This provides compatibility with Python’s \(0\)-indexing.

sage: Composition(descents=[1,0,4,8,11])
[1, 1, 3, 4, 3]
sage: Composition(descents=[0,1,3,4])
[1, 1, 2, 1]
sage: Composition(descents=([0,1,3],5))
[1, 1, 2, 1]
sage: Composition(descents=({0,1,3},5))
[1, 1, 2, 1]

>>> from sage.all import *
>>> Composition(descents=[Integer(1),Integer(0),Integer(4),Integer(8),Integer(11)])
[1, 1, 3, 4, 3]
>>> Composition(descents=[Integer(0),Integer(1),Integer(3),Integer(4)])
[1, 1, 2, 1]
>>> Composition(descents=([Integer(0),Integer(1),Integer(3)],Integer(5)))
[1, 1, 2, 1]
>>> Composition(descents=({Integer(0),Integer(1),Integer(3)},Integer(5)))
[1, 1, 2, 1]

An integer composition may be regarded as a sequence. Thus it is an instance of the Python abstract base class Sequence allows us to check if objects behave “like” sequences based on implemented methods. Note that collections.abc.Sequence is not the same as sage.structure.sequence.Sequence:

sage: import collections.abc
sage: C = Composition([3,2,3])
sage: isinstance(C, collections.abc.Sequence)
True
sage: issubclass(C.__class__, collections.abc.Sequence)
True

>>> from sage.all import *
>>> import collections.abc
>>> C = Composition([Integer(3),Integer(2),Integer(3)])
>>> isinstance(C, collections.abc.Sequence)
True
>>> issubclass(C.__class__, collections.abc.Sequence)
True

Typically, instances of collections.abc.Sequence have a .count method. Composition.count counts the number of parts of a specified size:

sage: C.count(3)
2

>>> from sage.all import *
>>> C.count(Integer(3))
2

EXAMPLES:

sage: C = Composition([3,1,2])
sage: TestSuite(C).run()

>>> from sage.all import *
>>> C = Composition([Integer(3),Integer(1),Integer(2)])
>>> TestSuite(C).run()
complement()[source]

Return the complement of the composition self.

The complement of a composition \(I\) is defined as follows:

If \(I\) is the empty composition, then the complement is the empty composition as well. Otherwise, let \(S\) be the descent set of \(I\) (that is, the subset \(\{ i_1, i_1 + i_2, \ldots, i_1 + i_2 + \cdots + i_{k-1} \}\) of \(\{ 1, 2, \ldots, |I|-1 \}\), where \(I\) is written as \((i_1, i_2, \ldots, i_k)\)). Then, the complement of \(I\) is defined as the composition of size \(|I|\) whose descent set is \(\{ 1, 2, \ldots, |I|-1 \} \setminus S\).

The complement of a composition \(I\) also is the reverse composition (reversed()) of the conjugate (conjugate()) of \(I\).

EXAMPLES:

sage: Composition([1, 1, 3, 1, 2, 1, 3]).conjugate()
[1, 1, 3, 3, 1, 3]
sage: Composition([1, 1, 3, 1, 2, 1, 3]).complement()
[3, 1, 3, 3, 1, 1]

>>> from sage.all import *
>>> Composition([Integer(1), Integer(1), Integer(3), Integer(1), Integer(2), Integer(1), Integer(3)]).conjugate()
[1, 1, 3, 3, 1, 3]
>>> Composition([Integer(1), Integer(1), Integer(3), Integer(1), Integer(2), Integer(1), Integer(3)]).complement()
[3, 1, 3, 3, 1, 1]
conjugate()[source]

Return the conjugate of the composition self.

The conjugate of a composition \(I\) is defined as the complement (see complement()) of the reverse composition (see reversed()) of \(I\).

An equivalent definition of the conjugate goes by saying that the ribbon shape of the conjugate of a composition \(I\) is the conjugate of the ribbon shape of \(I\). (The ribbon shape of a composition is returned by to_skew_partition().)

This implementation uses the algorithm from mupad-combinat.

EXAMPLES:

sage: Composition([1, 1, 3, 1, 2, 1, 3]).conjugate()
[1, 1, 3, 3, 1, 3]

>>> from sage.all import *
>>> Composition([Integer(1), Integer(1), Integer(3), Integer(1), Integer(2), Integer(1), Integer(3)]).conjugate()
[1, 1, 3, 3, 1, 3]

The ribbon shape of the conjugate of \(I\) is the conjugate of the ribbon shape of \(I\):

sage: all( I.conjugate().to_skew_partition()                                # needs sage.combinat
....:      == I.to_skew_partition().conjugate()
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( I.conjugate().to_skew_partition()                                # needs sage.combinat
...      == I.to_skew_partition().conjugate()
...      for I in Compositions(Integer(4)) )
True
count(n)[source]

Return the number of parts of size n.

EXAMPLES:

sage: C = Composition([3,2,3])
sage: C.count(3)
2
sage: C.count(2)
1
sage: C.count(1)
0

>>> from sage.all import *
>>> C = Composition([Integer(3),Integer(2),Integer(3)])
>>> C.count(Integer(3))
2
>>> C.count(Integer(2))
1
>>> C.count(Integer(1))
0
descents(final_descent=False)[source]

This gives one fewer than the partial sums of the composition.

This is here to maintain some sort of backward compatibility, even through the original implementation was broken (it gave the wrong answer). The same information can be found in partial_sums().

See also

partial_sums()

INPUT:

  • final_descent – boolean (default: False)

OUTPUT:

  • the list of partial sums of self with each part decremented by \(1\). This includes the sum of all entries when final_descent is True.

EXAMPLES:

sage: c = Composition([2,1,3,2])
sage: c.descents()
[1, 2, 5]
sage: c.descents(final_descent=True)
[1, 2, 5, 7]

>>> from sage.all import *
>>> c = Composition([Integer(2),Integer(1),Integer(3),Integer(2)])
>>> c.descents()
[1, 2, 5]
>>> c.descents(final_descent=True)
[1, 2, 5, 7]
fatten(grouping)[source]

Return the composition fatter than self, obtained by grouping together consecutive parts according to grouping.

INPUT:

  • grouping – a composition whose sum is the length of self

EXAMPLES:

Let us start with the composition:

sage: c = Composition([4,5,2,7,1])

>>> from sage.all import *
>>> c = Composition([Integer(4),Integer(5),Integer(2),Integer(7),Integer(1)])

With grouping equal to \((1, \ldots, 1)\), \(c\) is left unchanged:

sage: c.fatten(Composition([1,1,1,1,1]))
[4, 5, 2, 7, 1]

>>> from sage.all import *
>>> c.fatten(Composition([Integer(1),Integer(1),Integer(1),Integer(1),Integer(1)]))
[4, 5, 2, 7, 1]

With grouping equal to \((\ell)\) where \(\ell\) is the length of \(c\), this yields the coarsest composition above \(c\):

sage: c.fatten(Composition([5]))
[19]

>>> from sage.all import *
>>> c.fatten(Composition([Integer(5)]))
[19]

Other values for grouping yield (all the) other compositions coarser than \(c\):

sage: c.fatten(Composition([2,1,2]))
[9, 2, 8]
sage: c.fatten(Composition([3,1,1]))
[11, 7, 1]

>>> from sage.all import *
>>> c.fatten(Composition([Integer(2),Integer(1),Integer(2)]))
[9, 2, 8]
>>> c.fatten(Composition([Integer(3),Integer(1),Integer(1)]))
[11, 7, 1]
fatter()[source]

Return the set of compositions which are fatter than self.

Complexity for generation: \(O(|c|)\) memory, \(O(|r|)\) time where \(|c|\) is the size of self and \(r\) is the result.

EXAMPLES:

sage: C = Composition([4,5,2]).fatter()
sage: C.cardinality()
4
sage: list(C)
[[4, 5, 2], [4, 7], [9, 2], [11]]

>>> from sage.all import *
>>> C = Composition([Integer(4),Integer(5),Integer(2)]).fatter()
>>> C.cardinality()
4
>>> list(C)
[[4, 5, 2], [4, 7], [9, 2], [11]]

Some extreme cases:

sage: list(Composition([5]).fatter())
[[5]]
sage: list(Composition([]).fatter())
[[]]
sage: list(Composition([1,1,1,1]).fatter()) == list(Compositions(4))
True

>>> from sage.all import *
>>> list(Composition([Integer(5)]).fatter())
[[5]]
>>> list(Composition([]).fatter())
[[]]
>>> list(Composition([Integer(1),Integer(1),Integer(1),Integer(1)]).fatter()) == list(Compositions(Integer(4)))
True
finer()[source]

Return the set of compositions which are finer than self.

EXAMPLES:

sage: C = Composition([3,2]).finer()
sage: C.cardinality()
8
sage: C.list()
[[1, 1, 1, 1, 1], [1, 1, 1, 2], [1, 2, 1, 1], [1, 2, 2], [2, 1, 1, 1], [2, 1, 2], [3, 1, 1], [3, 2]]

sage: Composition([]).finer()
{[]}

>>> from sage.all import *
>>> C = Composition([Integer(3),Integer(2)]).finer()
>>> C.cardinality()
8
>>> C.list()
[[1, 1, 1, 1, 1], [1, 1, 1, 2], [1, 2, 1, 1], [1, 2, 2], [2, 1, 1, 1], [2, 1, 2], [3, 1, 1], [3, 2]]

>>> Composition([]).finer()
{[]}
inf(other, check=True)[source]

Return the meet of self with a composition other of the same size.

The meet of two compositions \(I\) and \(J\) of size \(n\) is the finest composition of \(n\) which is coarser than each of \(I\) and \(J\). It can be described as the composition whose descent set is the intersection of the descent sets of \(I\) and \(J\).

INPUT:

  • other – composition of same size as self

  • check – boolean (default: True); whether to check the input compositions for having the same size

OUTPUT: the meet of the compositions self and other

EXAMPLES:

sage: Composition([3, 1, 1, 3, 1]).meet([4, 3, 2])
[4, 5]
sage: Composition([9, 6]).meet([1, 3, 6, 3, 2])
[15]
sage: Composition([9, 6]).meet([1, 3, 5, 1, 3, 2])
[9, 6]
sage: Composition([1, 1, 1, 1, 1]).meet([3, 2])
[3, 2]
sage: Composition([4, 2]).meet([3, 3])
[6]
sage: Composition([]).meet([])
[]
sage: Composition([1]).meet([1])
[1]

>>> from sage.all import *
>>> Composition([Integer(3), Integer(1), Integer(1), Integer(3), Integer(1)]).meet([Integer(4), Integer(3), Integer(2)])
[4, 5]
>>> Composition([Integer(9), Integer(6)]).meet([Integer(1), Integer(3), Integer(6), Integer(3), Integer(2)])
[15]
>>> Composition([Integer(9), Integer(6)]).meet([Integer(1), Integer(3), Integer(5), Integer(1), Integer(3), Integer(2)])
[9, 6]
>>> Composition([Integer(1), Integer(1), Integer(1), Integer(1), Integer(1)]).meet([Integer(3), Integer(2)])
[3, 2]
>>> Composition([Integer(4), Integer(2)]).meet([Integer(3), Integer(3)])
[6]
>>> Composition([]).meet([])
[]
>>> Composition([Integer(1)]).meet([Integer(1)])
[1]

Let us verify on small examples that the meet of \(I\) and \(J\) is coarser than both of \(I\) and \(J\):

sage: all( all( I.is_finer(I.meet(J)) and
....:           J.is_finer(I.meet(J))
....:           for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( I.is_finer(I.meet(J)) and
...           J.is_finer(I.meet(J))
...           for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

and is the finest composition to do so:

sage: all( all( all( I.meet(J).is_finer(K)
....:                for K in I.fatter()
....:                if J.is_finer(K) )
....:           for J in Compositions(3) )
....:      for I in Compositions(3) )
True

>>> from sage.all import *
>>> all( all( all( I.meet(J).is_finer(K)
...                for K in I.fatter()
...                if J.is_finer(K) )
...           for J in Compositions(Integer(3)) )
...      for I in Compositions(Integer(3)) )
True

The descent set of the meet of \(I\) and \(J\) is the intersection of the descent sets of \(I\) and \(J\):

sage: def test_meet(n):
....:     return all( all( I.to_subset().intersection(J.to_subset())
....:                      == I.meet(J).to_subset()
....:                      for J in Compositions(n) )
....:                 for I in Compositions(n) )
sage: all( test_meet(n) for n in range(1, 5) )
True

>>> from sage.all import *
>>> def test_meet(n):
...     return all( all( I.to_subset().intersection(J.to_subset())
...                      == I.meet(J).to_subset()
...                      for J in Compositions(n) )
...                 for I in Compositions(n) )
>>> all( test_meet(n) for n in range(Integer(1), Integer(5)) )
True

See also

join()

AUTHORS:

  • Darij Grinberg (2013-09-05)

is_finer(co2)[source]

Return True if the composition self is finer than the composition co2; otherwise, return False.

EXAMPLES:

sage: Composition([4,1,2]).is_finer([3,1,3])
False
sage: Composition([3,1,3]).is_finer([4,1,2])
False
sage: Composition([1,2,2,1,1,2]).is_finer([5,1,3])
True
sage: Composition([2,2,2]).is_finer([4,2])
True

>>> from sage.all import *
>>> Composition([Integer(4),Integer(1),Integer(2)]).is_finer([Integer(3),Integer(1),Integer(3)])
False
>>> Composition([Integer(3),Integer(1),Integer(3)]).is_finer([Integer(4),Integer(1),Integer(2)])
False
>>> Composition([Integer(1),Integer(2),Integer(2),Integer(1),Integer(1),Integer(2)]).is_finer([Integer(5),Integer(1),Integer(3)])
True
>>> Composition([Integer(2),Integer(2),Integer(2)]).is_finer([Integer(4),Integer(2)])
True
join(other, check=True)[source]

Return the join of self with a composition other of the same size.

The join of two compositions \(I\) and \(J\) of size \(n\) is the coarsest composition of \(n\) which refines each of \(I\) and \(J\). It can be described as the composition whose descent set is the union of the descent sets of \(I\) and \(J\). It is also the concatenation of \(I_1, I_2, \cdots , I_m\), where \(I = I_1 \bullet I_2 \bullet \ldots \bullet I_m\) is the ribbon decomposition of \(I\) with respect to \(J\) (see ribbon_decomposition()).

INPUT:

  • other – composition of same size as self

  • check – boolean (default: True); whether to check the input compositions for having the same size

OUTPUT: the join of the compositions self and other

EXAMPLES:

sage: Composition([3, 1, 1, 3, 1]).join([4, 3, 2])
[3, 1, 1, 2, 1, 1]
sage: Composition([9, 6]).join([1, 3, 6, 3, 2])
[1, 3, 5, 1, 3, 2]
sage: Composition([9, 6]).join([1, 3, 5, 1, 3, 2])
[1, 3, 5, 1, 3, 2]
sage: Composition([1, 1, 1, 1, 1]).join([3, 2])
[1, 1, 1, 1, 1]
sage: Composition([4, 2]).join([3, 3])
[3, 1, 2]
sage: Composition([]).join([])
[]

>>> from sage.all import *
>>> Composition([Integer(3), Integer(1), Integer(1), Integer(3), Integer(1)]).join([Integer(4), Integer(3), Integer(2)])
[3, 1, 1, 2, 1, 1]
>>> Composition([Integer(9), Integer(6)]).join([Integer(1), Integer(3), Integer(6), Integer(3), Integer(2)])
[1, 3, 5, 1, 3, 2]
>>> Composition([Integer(9), Integer(6)]).join([Integer(1), Integer(3), Integer(5), Integer(1), Integer(3), Integer(2)])
[1, 3, 5, 1, 3, 2]
>>> Composition([Integer(1), Integer(1), Integer(1), Integer(1), Integer(1)]).join([Integer(3), Integer(2)])
[1, 1, 1, 1, 1]
>>> Composition([Integer(4), Integer(2)]).join([Integer(3), Integer(3)])
[3, 1, 2]
>>> Composition([]).join([])
[]

Let us verify on small examples that the join of \(I\) and \(J\) refines both of \(I\) and \(J\):

sage: all( all( I.join(J).is_finer(I) and
....:           I.join(J).is_finer(J)
....:           for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( I.join(J).is_finer(I) and
...           I.join(J).is_finer(J)
...           for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

and is the coarsest composition to do so:

sage: all( all( all( K.is_finer(I.join(J))
....:                for K in I.finer()
....:                if K.is_finer(J) )
....:           for J in Compositions(3) )
....:      for I in Compositions(3) )
True

>>> from sage.all import *
>>> all( all( all( K.is_finer(I.join(J))
...                for K in I.finer()
...                if K.is_finer(J) )
...           for J in Compositions(Integer(3)) )
...      for I in Compositions(Integer(3)) )
True

Let us check that the join of \(I\) and \(J\) is indeed the concatenation of \(I_1, I_2, \cdots , I_m\), where \(I = I_1 \bullet I_2 \bullet \ldots \bullet I_m\) is the ribbon decomposition of \(I\) with respect to \(J\):

sage: all( all( Composition.sum(I.ribbon_decomposition(J)[0])
....:           == I.join(J) for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( Composition.sum(I.ribbon_decomposition(J)[Integer(0)])
...           == I.join(J) for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

Also, the descent set of the join of \(I\) and \(J\) is the union of the descent sets of \(I\) and \(J\):

sage: all( all( I.to_subset().union(J.to_subset())
....:           == I.join(J).to_subset()
....:           for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( I.to_subset().union(J.to_subset())
...           == I.join(J).to_subset()
...           for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

See also

meet(), ribbon_decomposition()

AUTHORS:

  • Darij Grinberg (2013-09-05)

major_index()[source]

Return the major index of self. The major index is defined as the sum of the descents.

EXAMPLES:

sage: Composition([1, 1, 3, 1, 2, 1, 3]).major_index()
31

>>> from sage.all import *
>>> Composition([Integer(1), Integer(1), Integer(3), Integer(1), Integer(2), Integer(1), Integer(3)]).major_index()
31
meet(other, check=True)[source]

Return the meet of self with a composition other of the same size.

The meet of two compositions \(I\) and \(J\) of size \(n\) is the finest composition of \(n\) which is coarser than each of \(I\) and \(J\). It can be described as the composition whose descent set is the intersection of the descent sets of \(I\) and \(J\).

INPUT:

  • other – composition of same size as self

  • check – boolean (default: True); whether to check the input compositions for having the same size

OUTPUT: the meet of the compositions self and other

EXAMPLES:

sage: Composition([3, 1, 1, 3, 1]).meet([4, 3, 2])
[4, 5]
sage: Composition([9, 6]).meet([1, 3, 6, 3, 2])
[15]
sage: Composition([9, 6]).meet([1, 3, 5, 1, 3, 2])
[9, 6]
sage: Composition([1, 1, 1, 1, 1]).meet([3, 2])
[3, 2]
sage: Composition([4, 2]).meet([3, 3])
[6]
sage: Composition([]).meet([])
[]
sage: Composition([1]).meet([1])
[1]

>>> from sage.all import *
>>> Composition([Integer(3), Integer(1), Integer(1), Integer(3), Integer(1)]).meet([Integer(4), Integer(3), Integer(2)])
[4, 5]
>>> Composition([Integer(9), Integer(6)]).meet([Integer(1), Integer(3), Integer(6), Integer(3), Integer(2)])
[15]
>>> Composition([Integer(9), Integer(6)]).meet([Integer(1), Integer(3), Integer(5), Integer(1), Integer(3), Integer(2)])
[9, 6]
>>> Composition([Integer(1), Integer(1), Integer(1), Integer(1), Integer(1)]).meet([Integer(3), Integer(2)])
[3, 2]
>>> Composition([Integer(4), Integer(2)]).meet([Integer(3), Integer(3)])
[6]
>>> Composition([]).meet([])
[]
>>> Composition([Integer(1)]).meet([Integer(1)])
[1]

Let us verify on small examples that the meet of \(I\) and \(J\) is coarser than both of \(I\) and \(J\):

sage: all( all( I.is_finer(I.meet(J)) and
....:           J.is_finer(I.meet(J))
....:           for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( I.is_finer(I.meet(J)) and
...           J.is_finer(I.meet(J))
...           for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

and is the finest composition to do so:

sage: all( all( all( I.meet(J).is_finer(K)
....:                for K in I.fatter()
....:                if J.is_finer(K) )
....:           for J in Compositions(3) )
....:      for I in Compositions(3) )
True

>>> from sage.all import *
>>> all( all( all( I.meet(J).is_finer(K)
...                for K in I.fatter()
...                if J.is_finer(K) )
...           for J in Compositions(Integer(3)) )
...      for I in Compositions(Integer(3)) )
True

The descent set of the meet of \(I\) and \(J\) is the intersection of the descent sets of \(I\) and \(J\):

sage: def test_meet(n):
....:     return all( all( I.to_subset().intersection(J.to_subset())
....:                      == I.meet(J).to_subset()
....:                      for J in Compositions(n) )
....:                 for I in Compositions(n) )
sage: all( test_meet(n) for n in range(1, 5) )
True

>>> from sage.all import *
>>> def test_meet(n):
...     return all( all( I.to_subset().intersection(J.to_subset())
...                      == I.meet(J).to_subset()
...                      for J in Compositions(n) )
...                 for I in Compositions(n) )
>>> all( test_meet(n) for n in range(Integer(1), Integer(5)) )
True

See also

join()

AUTHORS:

  • Darij Grinberg (2013-09-05)

near_concatenation(other)[source]

Return the near-concatenation of two nonempty compositions self and other.

The near-concatenation \(I \odot J\) of two nonempty compositions \(I\) and \(J\) is defined as the composition \((i_1, i_2, \ldots , i_{n-1}, i_n + j_1, j_2, j_3, \ldots , j_m)\), where \((i_1, i_2, \ldots , i_n) = I\) and \((j_1, j_2, \ldots , j_m) = J\).

This method returns None if one of the two input compositions is empty.

EXAMPLES:

sage: Composition([1, 1, 3]).near_concatenation(Composition([4, 1, 2]))
[1, 1, 7, 1, 2]
sage: Composition([6]).near_concatenation(Composition([1, 5]))
[7, 5]
sage: Composition([1, 5]).near_concatenation(Composition([6]))
[1, 11]

>>> from sage.all import *
>>> Composition([Integer(1), Integer(1), Integer(3)]).near_concatenation(Composition([Integer(4), Integer(1), Integer(2)]))
[1, 1, 7, 1, 2]
>>> Composition([Integer(6)]).near_concatenation(Composition([Integer(1), Integer(5)]))
[7, 5]
>>> Composition([Integer(1), Integer(5)]).near_concatenation(Composition([Integer(6)]))
[1, 11]
partial_sums(final=True)[source]

The partial sums of the sequence defined by the entries of the composition.

If \(I = (i_1, \ldots, i_m)\) is a composition, then the partial sums of the entries of the composition are \([i_1, i_1 + i_2, \ldots, i_1 + i_2 + \cdots + i_m]\).

INPUT:

  • final – boolean (default: True); whether or not to include the final partial sum, which is always the size of the composition

See also

to_subset()

EXAMPLES:

sage: Composition([1,1,3,1,2,1,3]).partial_sums()
[1, 2, 5, 6, 8, 9, 12]

>>> from sage.all import *
>>> Composition([Integer(1),Integer(1),Integer(3),Integer(1),Integer(2),Integer(1),Integer(3)]).partial_sums()
[1, 2, 5, 6, 8, 9, 12]

With final = False, the last partial sum is not included:

sage: Composition([1,1,3,1,2,1,3]).partial_sums(final=False)
[1, 2, 5, 6, 8, 9]

>>> from sage.all import *
>>> Composition([Integer(1),Integer(1),Integer(3),Integer(1),Integer(2),Integer(1),Integer(3)]).partial_sums(final=False)
[1, 2, 5, 6, 8, 9]
peaks()[source]

Return a list of the peaks of the composition self.

The peaks of a composition are the descents which do not immediately follow another descent.

EXAMPLES:

sage: Composition([1, 1, 3, 1, 2, 1, 3]).peaks()
[4, 7]

>>> from sage.all import *
>>> Composition([Integer(1), Integer(1), Integer(3), Integer(1), Integer(2), Integer(1), Integer(3)]).peaks()
[4, 7]
refinement_splitting(J)[source]

Return the refinement splitting of self according to J.

INPUT:

  • J – a composition such that self is finer than J

OUTPUT:

  • the unique list of compositions \((I^{(p)})_{p=1, \ldots , m}\), obtained by splitting \(I\), such that \(|I^{(p)}| = J_p\) for all \(p = 1, \ldots, m\).

See also

refinement_splitting_lengths()

EXAMPLES:

sage: Composition([1,2,2,1,1,2]).refinement_splitting([5,1,3])
[[1, 2, 2], [1], [1, 2]]
sage: Composition([]).refinement_splitting([])
[]
sage: Composition([3]).refinement_splitting([2])
Traceback (most recent call last):
...
ValueError: compositions self (= [3]) and J (= [2]) must be of the same size
sage: Composition([2,1]).refinement_splitting([1,2])
Traceback (most recent call last):
...
ValueError: composition J (= [2, 1]) does not refine self (= [1, 2])

>>> from sage.all import *
>>> Composition([Integer(1),Integer(2),Integer(2),Integer(1),Integer(1),Integer(2)]).refinement_splitting([Integer(5),Integer(1),Integer(3)])
[[1, 2, 2], [1], [1, 2]]
>>> Composition([]).refinement_splitting([])
[]
>>> Composition([Integer(3)]).refinement_splitting([Integer(2)])
Traceback (most recent call last):
...
ValueError: compositions self (= [3]) and J (= [2]) must be of the same size
>>> Composition([Integer(2),Integer(1)]).refinement_splitting([Integer(1),Integer(2)])
Traceback (most recent call last):
...
ValueError: composition J (= [2, 1]) does not refine self (= [1, 2])
refinement_splitting_lengths(J)[source]

Return the lengths of the compositions in the refinement splitting of self according to J.

See also

refinement_splitting() for the definition of refinement splitting

EXAMPLES:

sage: Composition([1,2,2,1,1,2]).refinement_splitting_lengths([5,1,3])
[3, 1, 2]
sage: Composition([]).refinement_splitting_lengths([])
[]
sage: Composition([3]).refinement_splitting_lengths([2])
Traceback (most recent call last):
...
ValueError: compositions self (= [3]) and J (= [2]) must be of the same size
sage: Composition([2,1]).refinement_splitting_lengths([1,2])
Traceback (most recent call last):
...
ValueError: composition J (= [2, 1]) does not refine self (= [1, 2])

>>> from sage.all import *
>>> Composition([Integer(1),Integer(2),Integer(2),Integer(1),Integer(1),Integer(2)]).refinement_splitting_lengths([Integer(5),Integer(1),Integer(3)])
[3, 1, 2]
>>> Composition([]).refinement_splitting_lengths([])
[]
>>> Composition([Integer(3)]).refinement_splitting_lengths([Integer(2)])
Traceback (most recent call last):
...
ValueError: compositions self (= [3]) and J (= [2]) must be of the same size
>>> Composition([Integer(2),Integer(1)]).refinement_splitting_lengths([Integer(1),Integer(2)])
Traceback (most recent call last):
...
ValueError: composition J (= [2, 1]) does not refine self (= [1, 2])
reversed()[source]

Return the reverse composition of self.

The reverse composition of a composition \((i_1, i_2, \ldots, i_k)\) is defined as the composition \((i_k, i_{k-1}, \ldots, i_1)\).

EXAMPLES:

sage: Composition([1, 1, 3, 1, 2, 1, 3]).reversed()
[3, 1, 2, 1, 3, 1, 1]

>>> from sage.all import *
>>> Composition([Integer(1), Integer(1), Integer(3), Integer(1), Integer(2), Integer(1), Integer(3)]).reversed()
[3, 1, 2, 1, 3, 1, 1]
ribbon_decomposition(other, check=True)[source]

Return a pair describing the ribbon decomposition of a composition self with respect to a composition other of the same size.

If \(I\) and \(J\) are two compositions of the same nonzero size, then the ribbon decomposition of \(I\) with respect to \(J\) is defined as follows: Write \(I\) and \(J\) as \(I = (i_1, i_2, \ldots , i_n)\) and \(J = (j_1, j_2, \ldots , j_m)\). Then, the equality \(I = I_1 \bullet I_2 \bullet \ldots \bullet I_m\) holds for a unique \(m\)-tuple \((I_1, I_2, \ldots , I_m)\) of compositions such that each \(I_k\) has size \(j_k\) and for a unique choice of \(m-1\) signs \(\bullet\) each of which is either the concatenation sign \(\cdot\) or the near-concatenation sign \(\odot\) (see __add__() and near_concatenation() for the definitions of these two signs). This \(m\)-tuple and this choice of signs together are said to form the ribbon decomposition of \(I\) with respect to \(J\). If \(I\) and \(J\) are empty, then the same definition applies, except that there are \(0\) rather than \(m-1\) signs.

See Section 4.8 of [NCSF1].

INPUT:

  • other – composition of same size as self

  • check – boolean (default: True); whether to check the input compositions for having the same size

OUTPUT:

  • a pair (u, v), where u is a tuple of compositions (corresponding to the \(m\)-tuple \((I_1, I_2, \ldots , I_m)\) in the above definition), and v is a tuple of \(0`s and `1`s (encoding the choice of signs `\bullet\) in the above definition, with a \(0\) standing for \(\cdot\) and a \(1\) standing for \(\odot\)).

EXAMPLES:

sage: Composition([3, 1, 1, 3, 1]).ribbon_decomposition([4, 3, 2])
(([3, 1], [1, 2], [1, 1]), (0, 1))
sage: Composition([9, 6]).ribbon_decomposition([1, 3, 6, 3, 2])
(([1], [3], [5, 1], [3], [2]), (1, 1, 1, 1))
sage: Composition([9, 6]).ribbon_decomposition([1, 3, 5, 1, 3, 2])
(([1], [3], [5], [1], [3], [2]), (1, 1, 0, 1, 1))
sage: Composition([1, 1, 1, 1, 1]).ribbon_decomposition([3, 2])
(([1, 1, 1], [1, 1]), (0,))
sage: Composition([4, 2]).ribbon_decomposition([6])
(([4, 2],), ())
sage: Composition([]).ribbon_decomposition([])
((), ())

>>> from sage.all import *
>>> Composition([Integer(3), Integer(1), Integer(1), Integer(3), Integer(1)]).ribbon_decomposition([Integer(4), Integer(3), Integer(2)])
(([3, 1], [1, 2], [1, 1]), (0, 1))
>>> Composition([Integer(9), Integer(6)]).ribbon_decomposition([Integer(1), Integer(3), Integer(6), Integer(3), Integer(2)])
(([1], [3], [5, 1], [3], [2]), (1, 1, 1, 1))
>>> Composition([Integer(9), Integer(6)]).ribbon_decomposition([Integer(1), Integer(3), Integer(5), Integer(1), Integer(3), Integer(2)])
(([1], [3], [5], [1], [3], [2]), (1, 1, 0, 1, 1))
>>> Composition([Integer(1), Integer(1), Integer(1), Integer(1), Integer(1)]).ribbon_decomposition([Integer(3), Integer(2)])
(([1, 1, 1], [1, 1]), (0,))
>>> Composition([Integer(4), Integer(2)]).ribbon_decomposition([Integer(6)])
(([4, 2],), ())
>>> Composition([]).ribbon_decomposition([])
((), ())

Let us check that the defining property \(I = I_1 \bullet I_2 \bullet \ldots \bullet I_m\) is satisfied:

sage: def compose_back(u, v):
....:     comp = u[0]
....:     r = len(v)
....:     if len(u) != r + 1:
....:         raise ValueError("something is wrong")
....:     for i in range(r):
....:         if v[i] == 0:
....:             comp += u[i + 1]
....:         else:
....:             comp = comp.near_concatenation(u[i + 1])
....:     return comp
sage: all( all( all( compose_back(*(I.ribbon_decomposition(J))) == I
....:                for J in Compositions(n) )
....:           for I in Compositions(n) )
....:      for n in range(1, 5) )
True

>>> from sage.all import *
>>> def compose_back(u, v):
...     comp = u[Integer(0)]
...     r = len(v)
...     if len(u) != r + Integer(1):
...         raise ValueError("something is wrong")
...     for i in range(r):
...         if v[i] == Integer(0):
...             comp += u[i + Integer(1)]
...         else:
...             comp = comp.near_concatenation(u[i + Integer(1)])
...     return comp
>>> all( all( all( compose_back(*(I.ribbon_decomposition(J))) == I
...                for J in Compositions(n) )
...           for I in Compositions(n) )
...      for n in range(Integer(1), Integer(5)) )
True

AUTHORS:

  • Darij Grinberg (2013-08-29)

shuffle_product(other, overlap=False)[source]

The (overlapping) shuffles of self and other.

Suppose \(I = (i_1, \ldots, i_k)\) and \(J = (j_1, \ldots, j_l)\) are two compositions. A shuffle of \(I\) and \(J\) is a composition of length \(k + l\) that contains both \(I\) and \(J\) as subsequences.

More generally, an overlapping shuffle of \(I\) and \(J\) is obtained by distributing the elements of \(I\) and \(J\) (preserving the relative ordering of these elements) among the positions of an empty list; an element of \(I\) and an element of \(J\) are permitted to share the same position, in which case they are replaced by their sum. In particular, a shuffle of \(I\) and \(J\) is an overlapping shuffle of \(I\) and \(J\).

INPUT:

  • other – composition

  • overlap – boolean (default: False); if True, the overlapping shuffle product is returned

OUTPUT:

An enumerated set (allowing for multiplicities)

EXAMPLES:

The shuffle product of \([2,2]\) and \([1,1,3]\):

sage: alph = Composition([2,2])
sage: beta = Composition([1,1,3])
sage: S = alph.shuffle_product(beta); S                                     # needs sage.combinat
Shuffle product of [2, 2] and [1, 1, 3]
sage: S.list()                                                              # needs sage.combinat
[[2, 2, 1, 1, 3], [2, 1, 2, 1, 3], [2, 1, 1, 2, 3], [2, 1, 1, 3, 2],
 [1, 2, 2, 1, 3], [1, 2, 1, 2, 3], [1, 2, 1, 3, 2], [1, 1, 2, 2, 3],
 [1, 1, 2, 3, 2], [1, 1, 3, 2, 2]]

>>> from sage.all import *
>>> alph = Composition([Integer(2),Integer(2)])
>>> beta = Composition([Integer(1),Integer(1),Integer(3)])
>>> S = alph.shuffle_product(beta); S                                     # needs sage.combinat
Shuffle product of [2, 2] and [1, 1, 3]
>>> S.list()                                                              # needs sage.combinat
[[2, 2, 1, 1, 3], [2, 1, 2, 1, 3], [2, 1, 1, 2, 3], [2, 1, 1, 3, 2],
 [1, 2, 2, 1, 3], [1, 2, 1, 2, 3], [1, 2, 1, 3, 2], [1, 1, 2, 2, 3],
 [1, 1, 2, 3, 2], [1, 1, 3, 2, 2]]

The overlapping shuffle product of \([2,2]\) and \([1,1,3]\):

sage: alph = Composition([2,2])
sage: beta = Composition([1,1,3])
sage: O = alph.shuffle_product(beta, overlap=True); O                       # needs sage.combinat
Overlapping shuffle product of [2, 2] and [1, 1, 3]
sage: O.list()                                                              # needs sage.combinat
[[2, 2, 1, 1, 3], [2, 1, 2, 1, 3], [2, 1, 1, 2, 3], [2, 1, 1, 3, 2],
 [1, 2, 2, 1, 3], [1, 2, 1, 2, 3], [1, 2, 1, 3, 2], [1, 1, 2, 2, 3],
 [1, 1, 2, 3, 2], [1, 1, 3, 2, 2],
 [3, 2, 1, 3], [2, 3, 1, 3], [3, 1, 2, 3], [2, 1, 3, 3], [3, 1, 3, 2],
 [2, 1, 1, 5], [1, 3, 2, 3], [1, 2, 3, 3], [1, 3, 3, 2], [1, 2, 1, 5],
 [1, 1, 5, 2], [1, 1, 2, 5],
 [3, 3, 3], [3, 1, 5], [1, 3, 5]]

>>> from sage.all import *
>>> alph = Composition([Integer(2),Integer(2)])
>>> beta = Composition([Integer(1),Integer(1),Integer(3)])
>>> O = alph.shuffle_product(beta, overlap=True); O                       # needs sage.combinat
Overlapping shuffle product of [2, 2] and [1, 1, 3]
>>> O.list()                                                              # needs sage.combinat
[[2, 2, 1, 1, 3], [2, 1, 2, 1, 3], [2, 1, 1, 2, 3], [2, 1, 1, 3, 2],
 [1, 2, 2, 1, 3], [1, 2, 1, 2, 3], [1, 2, 1, 3, 2], [1, 1, 2, 2, 3],
 [1, 1, 2, 3, 2], [1, 1, 3, 2, 2],
 [3, 2, 1, 3], [2, 3, 1, 3], [3, 1, 2, 3], [2, 1, 3, 3], [3, 1, 3, 2],
 [2, 1, 1, 5], [1, 3, 2, 3], [1, 2, 3, 3], [1, 3, 3, 2], [1, 2, 1, 5],
 [1, 1, 5, 2], [1, 1, 2, 5],
 [3, 3, 3], [3, 1, 5], [1, 3, 5]]

Note that the shuffle product of two compositions can include the same composition more than once since a composition can be a shuffle of two compositions in several ways. For example:

sage: # needs sage.combinat
sage: w1 = Composition([1])
sage: S = w1.shuffle_product(w1); S
Shuffle product of [1] and [1]
sage: S.list()
[[1, 1], [1, 1]]
sage: O = w1.shuffle_product(w1, overlap=True); O
Overlapping shuffle product of [1] and [1]
sage: O.list()
[[1, 1], [1, 1], [2]]

>>> from sage.all import *
>>> # needs sage.combinat
>>> w1 = Composition([Integer(1)])
>>> S = w1.shuffle_product(w1); S
Shuffle product of [1] and [1]
>>> S.list()
[[1, 1], [1, 1]]
>>> O = w1.shuffle_product(w1, overlap=True); O
Overlapping shuffle product of [1] and [1]
>>> O.list()
[[1, 1], [1, 1], [2]]
size()[source]

Return the size of self, that is the sum of its parts.

EXAMPLES:

sage: Composition([7,1,3]).size()
11

>>> from sage.all import *
>>> Composition([Integer(7),Integer(1),Integer(3)]).size()
11
specht_module(base_ring=None)[source]

Return the Specht module corresponding to self.

EXAMPLES:

sage: SM = Composition([1,2,2]).specht_module(QQ); SM                       # needs sage.combinat sage.modules
Specht module of [(0, 0), (1, 0), (1, 1), (2, 0), (2, 1)] over Rational Field
sage: s = SymmetricFunctions(QQ).s()                                        # needs sage.combinat sage.modules
sage: s(SM.frobenius_image())                                               # needs sage.combinat sage.modules
s[2, 2, 1]

>>> from sage.all import *
>>> SM = Composition([Integer(1),Integer(2),Integer(2)]).specht_module(QQ); SM                       # needs sage.combinat sage.modules
Specht module of [(0, 0), (1, 0), (1, 1), (2, 0), (2, 1)] over Rational Field
>>> s = SymmetricFunctions(QQ).s()                                        # needs sage.combinat sage.modules
>>> s(SM.frobenius_image())                                               # needs sage.combinat sage.modules
s[2, 2, 1]
specht_module_dimension(base_ring=None)[source]

Return the dimension of the Specht module corresponding to self.

INPUT:

  • base_ring – (default: \(\QQ\)) the base ring

EXAMPLES:

sage: Composition([1,2,2]).specht_module_dimension()                        # needs sage.combinat sage.modules
5
sage: Composition([1,2,2]).specht_module_dimension(GF(2))                   # needs sage.combinat sage.modules sage.rings.finite_rings
5

>>> from sage.all import *
>>> Composition([Integer(1),Integer(2),Integer(2)]).specht_module_dimension()                        # needs sage.combinat sage.modules
5
>>> Composition([Integer(1),Integer(2),Integer(2)]).specht_module_dimension(GF(Integer(2)))                   # needs sage.combinat sage.modules sage.rings.finite_rings
5
static sum(compositions)[source]

Return the concatenation of the given compositions.

INPUT:

  • compositions – list (or iterable) of compositions

EXAMPLES:

sage: Composition.sum([Composition([1, 1, 3]), Composition([4, 1, 2]), Composition([3,1])])
[1, 1, 3, 4, 1, 2, 3, 1]

>>> from sage.all import *
>>> Composition.sum([Composition([Integer(1), Integer(1), Integer(3)]), Composition([Integer(4), Integer(1), Integer(2)]), Composition([Integer(3),Integer(1)])])
[1, 1, 3, 4, 1, 2, 3, 1]

Any iterable can be provided as input:

sage: Composition.sum([Composition([i,i]) for i in [4,1,3]])
[4, 4, 1, 1, 3, 3]

>>> from sage.all import *
>>> Composition.sum([Composition([i,i]) for i in [Integer(4),Integer(1),Integer(3)]])
[4, 4, 1, 1, 3, 3]

Empty inputs are handled gracefully:

sage: Composition.sum([]) == Composition([])
True

>>> from sage.all import *
>>> Composition.sum([]) == Composition([])
True
sup(other, check=True)[source]

Return the join of self with a composition other of the same size.

The join of two compositions \(I\) and \(J\) of size \(n\) is the coarsest composition of \(n\) which refines each of \(I\) and \(J\). It can be described as the composition whose descent set is the union of the descent sets of \(I\) and \(J\). It is also the concatenation of \(I_1, I_2, \cdots , I_m\), where \(I = I_1 \bullet I_2 \bullet \ldots \bullet I_m\) is the ribbon decomposition of \(I\) with respect to \(J\) (see ribbon_decomposition()).

INPUT:

  • other – composition of same size as self

  • check – boolean (default: True); whether to check the input compositions for having the same size

OUTPUT: the join of the compositions self and other

EXAMPLES:

sage: Composition([3, 1, 1, 3, 1]).join([4, 3, 2])
[3, 1, 1, 2, 1, 1]
sage: Composition([9, 6]).join([1, 3, 6, 3, 2])
[1, 3, 5, 1, 3, 2]
sage: Composition([9, 6]).join([1, 3, 5, 1, 3, 2])
[1, 3, 5, 1, 3, 2]
sage: Composition([1, 1, 1, 1, 1]).join([3, 2])
[1, 1, 1, 1, 1]
sage: Composition([4, 2]).join([3, 3])
[3, 1, 2]
sage: Composition([]).join([])
[]

>>> from sage.all import *
>>> Composition([Integer(3), Integer(1), Integer(1), Integer(3), Integer(1)]).join([Integer(4), Integer(3), Integer(2)])
[3, 1, 1, 2, 1, 1]
>>> Composition([Integer(9), Integer(6)]).join([Integer(1), Integer(3), Integer(6), Integer(3), Integer(2)])
[1, 3, 5, 1, 3, 2]
>>> Composition([Integer(9), Integer(6)]).join([Integer(1), Integer(3), Integer(5), Integer(1), Integer(3), Integer(2)])
[1, 3, 5, 1, 3, 2]
>>> Composition([Integer(1), Integer(1), Integer(1), Integer(1), Integer(1)]).join([Integer(3), Integer(2)])
[1, 1, 1, 1, 1]
>>> Composition([Integer(4), Integer(2)]).join([Integer(3), Integer(3)])
[3, 1, 2]
>>> Composition([]).join([])
[]

Let us verify on small examples that the join of \(I\) and \(J\) refines both of \(I\) and \(J\):

sage: all( all( I.join(J).is_finer(I) and
....:           I.join(J).is_finer(J)
....:           for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( I.join(J).is_finer(I) and
...           I.join(J).is_finer(J)
...           for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

and is the coarsest composition to do so:

sage: all( all( all( K.is_finer(I.join(J))
....:                for K in I.finer()
....:                if K.is_finer(J) )
....:           for J in Compositions(3) )
....:      for I in Compositions(3) )
True

>>> from sage.all import *
>>> all( all( all( K.is_finer(I.join(J))
...                for K in I.finer()
...                if K.is_finer(J) )
...           for J in Compositions(Integer(3)) )
...      for I in Compositions(Integer(3)) )
True

Let us check that the join of \(I\) and \(J\) is indeed the concatenation of \(I_1, I_2, \cdots , I_m\), where \(I = I_1 \bullet I_2 \bullet \ldots \bullet I_m\) is the ribbon decomposition of \(I\) with respect to \(J\):

sage: all( all( Composition.sum(I.ribbon_decomposition(J)[0])
....:           == I.join(J) for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( Composition.sum(I.ribbon_decomposition(J)[Integer(0)])
...           == I.join(J) for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

Also, the descent set of the join of \(I\) and \(J\) is the union of the descent sets of \(I\) and \(J\):

sage: all( all( I.to_subset().union(J.to_subset())
....:           == I.join(J).to_subset()
....:           for J in Compositions(4) )
....:      for I in Compositions(4) )
True

>>> from sage.all import *
>>> all( all( I.to_subset().union(J.to_subset())
...           == I.join(J).to_subset()
...           for J in Compositions(Integer(4)) )
...      for I in Compositions(Integer(4)) )
True

See also

meet(), ribbon_decomposition()

AUTHORS:

  • Darij Grinberg (2013-09-05)

to_code()[source]

Return the code of the composition self.

The code of a composition \(I\) is a list of length \(\mathrm{size}(I)\) of 1s and 0s such that there is a 1 wherever a new part starts. (Exceptional case: When the composition is empty, the code is [0].)

EXAMPLES:

sage: Composition([4,1,2,3,5]).to_code()
[1, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]

>>> from sage.all import *
>>> Composition([Integer(4),Integer(1),Integer(2),Integer(3),Integer(5)]).to_code()
[1, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
to_partition()[source]

Return the partition obtained by sorting self into decreasing order.

EXAMPLES:

sage: Composition([2,1,3]).to_partition()                                   # needs sage.combinat
[3, 2, 1]
sage: Composition([4,2,2]).to_partition()                                   # needs sage.combinat
[4, 2, 2]
sage: Composition([]).to_partition()                                        # needs sage.combinat
[]

>>> from sage.all import *
>>> Composition([Integer(2),Integer(1),Integer(3)]).to_partition()                                   # needs sage.combinat
[3, 2, 1]
>>> Composition([Integer(4),Integer(2),Integer(2)]).to_partition()                                   # needs sage.combinat
[4, 2, 2]
>>> Composition([]).to_partition()                                        # needs sage.combinat
[]
to_skew_partition(overlap=1)[source]

Return the skew partition obtained from self.

This is a skew partition whose rows have the entries of self as their length, taken in reverse order (so the first entry of self is the length of the lowermost row, etc.). The parameter overlap indicates the number of cells on each row that are directly below cells of the previous row. When it is set to \(1\) (its default value), the result is the ribbon shape of self.

EXAMPLES:

sage: # needs sage.combinat
sage: Composition([3,4,1]).to_skew_partition()
[6, 6, 3] / [5, 2]
sage: Composition([3,4,1]).to_skew_partition(overlap=0)
[8, 7, 3] / [7, 3]
sage: Composition([]).to_skew_partition()
[] / []
sage: Composition([1,2]).to_skew_partition()
[2, 1] / []
sage: Composition([2,1]).to_skew_partition()
[2, 2] / [1]

>>> from sage.all import *
>>> # needs sage.combinat
>>> Composition([Integer(3),Integer(4),Integer(1)]).to_skew_partition()
[6, 6, 3] / [5, 2]
>>> Composition([Integer(3),Integer(4),Integer(1)]).to_skew_partition(overlap=Integer(0))
[8, 7, 3] / [7, 3]
>>> Composition([]).to_skew_partition()
[] / []
>>> Composition([Integer(1),Integer(2)]).to_skew_partition()
[2, 1] / []
>>> Composition([Integer(2),Integer(1)]).to_skew_partition()
[2, 2] / [1]
to_subset(final=False)[source]

The subset corresponding to self under the bijection (see below) between compositions of \(n\) and subsets of \(\{1, 2, \ldots, n-1\}\).

The bijection maps a composition \((i_1, \ldots, i_k)\) of \(n\) to \(\{i_1, i_1 + i_2, i_1 + i_2 + i_3, \ldots, i_1 + \cdots + i_{k-1}\}\).

INPUT:

  • final – boolean (default: False); whether or not to include the final partial sum, which is always the size of the composition

See also

partial_sums()

EXAMPLES:

sage: Composition([1,1,3,1,2,1,3]).to_subset()
{1, 2, 5, 6, 8, 9}
sage: for I in Compositions(3): print(I.to_subset())
{1, 2}
{1}
{2}
{}

>>> from sage.all import *
>>> Composition([Integer(1),Integer(1),Integer(3),Integer(1),Integer(2),Integer(1),Integer(3)]).to_subset()
{1, 2, 5, 6, 8, 9}
>>> for I in Compositions(Integer(3)): print(I.to_subset())
{1, 2}
{1}
{2}
{}

With final=True, the sum of all the elements of the composition is included in the subset:

sage: Composition([1,1,3,1,2,1,3]).to_subset(final=True)
{1, 2, 5, 6, 8, 9, 12}

>>> from sage.all import *
>>> Composition([Integer(1),Integer(1),Integer(3),Integer(1),Integer(2),Integer(1),Integer(3)]).to_subset(final=True)
{1, 2, 5, 6, 8, 9, 12}
wll_gt(co2)[source]

Return True if the composition self is greater than the composition co2 with respect to the wll-ordering; otherwise, return False.

The wll-ordering is a total order on the set of all compositions defined as follows: A composition \(I\) is greater than a composition \(J\) if and only if one of the following conditions holds:

  • The size of \(I\) is greater than the size of \(J\).

  • The size of \(I\) equals the size of \(J\), but the length of \(I\) is greater than the length of \(J\).

  • The size of \(I\) equals the size of \(J\), and the length of \(I\) equals the length of \(J\), but \(I\) is lexicographically greater than \(J\).

(“wll-ordering” is short for “weight, length, lexicographic ordering”.)

EXAMPLES:

sage: Composition([4,1,2]).wll_gt([3,1,3])
True
sage: Composition([7]).wll_gt([4,1,2])
False
sage: Composition([8]).wll_gt([4,1,2])
True
sage: Composition([3,2,2,2]).wll_gt([5,2])
True
sage: Composition([]).wll_gt([3])
False
sage: Composition([2,1]).wll_gt([2,1])
False
sage: Composition([2,2,2]).wll_gt([4,2])
True
sage: Composition([4,2]).wll_gt([2,2,2])
False
sage: Composition([1,1,2]).wll_gt([2,2])
True
sage: Composition([2,2]).wll_gt([1,3])
True
sage: Composition([2,1,2]).wll_gt([])
True

>>> from sage.all import *
>>> Composition([Integer(4),Integer(1),Integer(2)]).wll_gt([Integer(3),Integer(1),Integer(3)])
True
>>> Composition([Integer(7)]).wll_gt([Integer(4),Integer(1),Integer(2)])
False
>>> Composition([Integer(8)]).wll_gt([Integer(4),Integer(1),Integer(2)])
True
>>> Composition([Integer(3),Integer(2),Integer(2),Integer(2)]).wll_gt([Integer(5),Integer(2)])
True
>>> Composition([]).wll_gt([Integer(3)])
False
>>> Composition([Integer(2),Integer(1)]).wll_gt([Integer(2),Integer(1)])
False
>>> Composition([Integer(2),Integer(2),Integer(2)]).wll_gt([Integer(4),Integer(2)])
True
>>> Composition([Integer(4),Integer(2)]).wll_gt([Integer(2),Integer(2),Integer(2)])
False
>>> Composition([Integer(1),Integer(1),Integer(2)]).wll_gt([Integer(2),Integer(2)])
True
>>> Composition([Integer(2),Integer(2)]).wll_gt([Integer(1),Integer(3)])
True
>>> Composition([Integer(2),Integer(1),Integer(2)]).wll_gt([])
True
class sage.combinat.composition.Compositions(is_infinite=False, category=None)[source]

Bases: UniqueRepresentation, Parent

Set of integer compositions.

A composition \(c\) of a nonnegative integer \(n\) is a list of positive integers with total sum \(n\).

See also

EXAMPLES:

There are 8 compositions of 4:

sage: Compositions(4).cardinality()
8

>>> from sage.all import *
>>> Compositions(Integer(4)).cardinality()
8

Here is the list of them:

sage: Compositions(4).list()
[[1, 1, 1, 1], [1, 1, 2], [1, 2, 1], [1, 3], [2, 1, 1], [2, 2], [3, 1], [4]]

>>> from sage.all import *
>>> Compositions(Integer(4)).list()
[[1, 1, 1, 1], [1, 1, 2], [1, 2, 1], [1, 3], [2, 1, 1], [2, 2], [3, 1], [4]]

You can use the .first() method to get the ‘first’ composition of a number:

sage: Compositions(4).first()
[1, 1, 1, 1]

>>> from sage.all import *
>>> Compositions(Integer(4)).first()
[1, 1, 1, 1]

You can also calculate the ‘next’ composition given the current one:

sage: Compositions(4).next([1,1,2])
[1, 2, 1]

>>> from sage.all import *
>>> Compositions(Integer(4)).next([Integer(1),Integer(1),Integer(2)])
[1, 2, 1]

If \(n\) is not specified, this returns the combinatorial class of all (nonnegative) integer compositions:

sage: Compositions()
Compositions of nonnegative integers
sage: [] in Compositions()
True
sage: [2,3,1] in Compositions()
True
sage: [-2,3,1] in Compositions()
False

>>> from sage.all import *
>>> Compositions()
Compositions of nonnegative integers
>>> [] in Compositions()
True
>>> [Integer(2),Integer(3),Integer(1)] in Compositions()
True
>>> [-Integer(2),Integer(3),Integer(1)] in Compositions()
False

If \(n\) is specified, it returns the class of compositions of \(n\):

sage: Compositions(3)
Compositions of 3
sage: list(Compositions(3))
[[1, 1, 1], [1, 2], [2, 1], [3]]
sage: Compositions(3).cardinality()
4

>>> from sage.all import *
>>> Compositions(Integer(3))
Compositions of 3
>>> list(Compositions(Integer(3)))
[[1, 1, 1], [1, 2], [2, 1], [3]]
>>> Compositions(Integer(3)).cardinality()
4

The following examples show how to test whether or not an object is a composition:

sage: [3,4] in Compositions()
True
sage: [3,4] in Compositions(7)
True
sage: [3,4] in Compositions(5)
False

>>> from sage.all import *
>>> [Integer(3),Integer(4)] in Compositions()
True
>>> [Integer(3),Integer(4)] in Compositions(Integer(7))
True
>>> [Integer(3),Integer(4)] in Compositions(Integer(5))
False

Similarly, one can check whether or not an object is a composition which satisfies further constraints:

sage: [4,2] in Compositions(6, inner=[2,2])
True
sage: [4,2] in Compositions(6, inner=[2,3])
False
sage: [4,1] in Compositions(5, inner=[2,1], max_slope = 0)
True

>>> from sage.all import *
>>> [Integer(4),Integer(2)] in Compositions(Integer(6), inner=[Integer(2),Integer(2)])
True
>>> [Integer(4),Integer(2)] in Compositions(Integer(6), inner=[Integer(2),Integer(3)])
False
>>> [Integer(4),Integer(1)] in Compositions(Integer(5), inner=[Integer(2),Integer(1)], max_slope = Integer(0))
True

An example with incompatible constraints:

sage: [4,2] in Compositions(6, inner=[2,2], min_part=3)
False

>>> from sage.all import *
>>> [Integer(4),Integer(2)] in Compositions(Integer(6), inner=[Integer(2),Integer(2)], min_part=Integer(3))
False

The options length, min_length, and max_length can be used to set length constraints on the compositions. For example, the compositions of 4 of length equal to, at least, and at most 2 are given by:

sage: Compositions(4, length=2).list()
[[3, 1], [2, 2], [1, 3]]
sage: Compositions(4, min_length=2).list()
[[3, 1], [2, 2], [2, 1, 1], [1, 3], [1, 2, 1], [1, 1, 2], [1, 1, 1, 1]]
sage: Compositions(4, max_length=2).list()
[[4], [3, 1], [2, 2], [1, 3]]

>>> from sage.all import *
>>> Compositions(Integer(4), length=Integer(2)).list()
[[3, 1], [2, 2], [1, 3]]
>>> Compositions(Integer(4), min_length=Integer(2)).list()
[[3, 1], [2, 2], [2, 1, 1], [1, 3], [1, 2, 1], [1, 1, 2], [1, 1, 1, 1]]
>>> Compositions(Integer(4), max_length=Integer(2)).list()
[[4], [3, 1], [2, 2], [1, 3]]

Setting both min_length and max_length to the same value is equivalent to setting length to this value:

sage: Compositions(4, min_length=2, max_length=2).list()
[[3, 1], [2, 2], [1, 3]]

>>> from sage.all import *
>>> Compositions(Integer(4), min_length=Integer(2), max_length=Integer(2)).list()
[[3, 1], [2, 2], [1, 3]]

The options inner and outer can be used to set part-by-part containment constraints. The list of compositions of 4 bounded above by [3,1,2] is given by:

sage: list(Compositions(4, outer=[3,1,2]))
[[3, 1], [2, 1, 1], [1, 1, 2]]

>>> from sage.all import *
>>> list(Compositions(Integer(4), outer=[Integer(3),Integer(1),Integer(2)]))
[[3, 1], [2, 1, 1], [1, 1, 2]]

outer sets max_length to the length of its argument. Moreover, the parts of outer may be infinite to clear the constraint on specific parts. This is the list of compositions of 4 of length at most 3 such that the first and third parts are at most 1:

sage: Compositions(4, outer=[1,oo,1]).list()
[[1, 3], [1, 2, 1]]

>>> from sage.all import *
>>> Compositions(Integer(4), outer=[Integer(1),oo,Integer(1)]).list()
[[1, 3], [1, 2, 1]]

This is the list of compositions of 4 bounded below by [1,1,1]:

sage: Compositions(4, inner=[1,1,1]).list()
[[2, 1, 1], [1, 2, 1], [1, 1, 2], [1, 1, 1, 1]]

>>> from sage.all import *
>>> Compositions(Integer(4), inner=[Integer(1),Integer(1),Integer(1)]).list()
[[2, 1, 1], [1, 2, 1], [1, 1, 2], [1, 1, 1, 1]]

The options min_slope and max_slope can be used to set constraints on the slope, that is the difference p[i+1]-p[i] of two consecutive parts. The following is the list of weakly increasing compositions of 4:

sage: Compositions(4, min_slope=0).list()
[[4], [2, 2], [1, 3], [1, 1, 2], [1, 1, 1, 1]]

>>> from sage.all import *
>>> Compositions(Integer(4), min_slope=Integer(0)).list()
[[4], [2, 2], [1, 3], [1, 1, 2], [1, 1, 1, 1]]

Here are the weakly decreasing ones:

sage: Compositions(4, max_slope=0).list()
[[4], [3, 1], [2, 2], [2, 1, 1], [1, 1, 1, 1]]

>>> from sage.all import *
>>> Compositions(Integer(4), max_slope=Integer(0)).list()
[[4], [3, 1], [2, 2], [2, 1, 1], [1, 1, 1, 1]]

The following is the list of compositions of 4 such that two consecutive parts differ by at most one:

sage: Compositions(4, min_slope=-1, max_slope=1).list()
[[4], [2, 2], [2, 1, 1], [1, 2, 1], [1, 1, 2], [1, 1, 1, 1]]

>>> from sage.all import *
>>> Compositions(Integer(4), min_slope=-Integer(1), max_slope=Integer(1)).list()
[[4], [2, 2], [2, 1, 1], [1, 2, 1], [1, 1, 2], [1, 1, 1, 1]]

The constraints can be combined together in all reasonable ways. This is the list of compositions of 5 of length between 2 and 4 such that the difference between consecutive parts is between -2 and 1:

sage: Compositions(5, max_slope=1, min_slope=-2, min_length=2, max_length=4).list()
[[3, 2], [3, 1, 1], [2, 3], [2, 2, 1], [2, 1, 2], [2, 1, 1, 1], [1, 2, 2], [1, 2, 1, 1], [1, 1, 2, 1], [1, 1, 1, 2]]

>>> from sage.all import *
>>> Compositions(Integer(5), max_slope=Integer(1), min_slope=-Integer(2), min_length=Integer(2), max_length=Integer(4)).list()
[[3, 2], [3, 1, 1], [2, 3], [2, 2, 1], [2, 1, 2], [2, 1, 1, 1], [1, 2, 2], [1, 2, 1, 1], [1, 1, 2, 1], [1, 1, 1, 2]]

We can do the same thing with an outer constraint:

sage: Compositions(5, max_slope=1, min_slope=-2, min_length=2, max_length=4, outer=[2,5,2]).list()
[[2, 3], [2, 2, 1], [2, 1, 2], [1, 2, 2]]

>>> from sage.all import *
>>> Compositions(Integer(5), max_slope=Integer(1), min_slope=-Integer(2), min_length=Integer(2), max_length=Integer(4), outer=[Integer(2),Integer(5),Integer(2)]).list()
[[2, 3], [2, 2, 1], [2, 1, 2], [1, 2, 2]]

However, providing incoherent constraints may yield strange results. It is up to the user to ensure that the inner and outer compositions themselves satisfy the parts and slope constraints.

Note that setting min_part=0 is not allowed:

sage: Compositions(2, length=3, min_part=0)
Traceback (most recent call last):
...
ValueError: setting min_part=0 is not allowed for Compositions

>>> from sage.all import *
>>> Compositions(Integer(2), length=Integer(3), min_part=Integer(0))
Traceback (most recent call last):
...
ValueError: setting min_part=0 is not allowed for Compositions

Instead you must use IntegerVectors:

sage: list(IntegerVectors(2, 3))
[[2, 0, 0], [1, 1, 0], [1, 0, 1], [0, 2, 0], [0, 1, 1], [0, 0, 2]]

>>> from sage.all import *
>>> list(IntegerVectors(Integer(2), Integer(3)))
[[2, 0, 0], [1, 1, 0], [1, 0, 1], [0, 2, 0], [0, 1, 1], [0, 0, 2]]

The generation algorithm is constant amortized time, and handled by the generic tool IntegerListsLex.

Element[source]

alias of Composition

from_code(code)[source]

Return the composition from its code. The code of a composition \(I\) is a list of length \(\mathrm{size}(I)\) consisting of 1s and 0s such that there is a 1 wherever a new part starts. (Exceptional case: When the composition is empty, the code is [0].)

EXAMPLES:

sage: Composition([4,1,2,3,5]).to_code()
[1, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
sage: Compositions().from_code(_)
[4, 1, 2, 3, 5]
sage: Composition([3,1,2,3,5]).to_code()
[1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
sage: Compositions().from_code(_)
[3, 1, 2, 3, 5]

>>> from sage.all import *
>>> Composition([Integer(4),Integer(1),Integer(2),Integer(3),Integer(5)]).to_code()
[1, 0, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
>>> Compositions().from_code(_)
[4, 1, 2, 3, 5]
>>> Composition([Integer(3),Integer(1),Integer(2),Integer(3),Integer(5)]).to_code()
[1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0]
>>> Compositions().from_code(_)
[3, 1, 2, 3, 5]
from_descents(descents, nps=None)[source]

Return a composition from the list of descents.

INPUT:

  • descents – an iterable

  • nps – integer or None (default: None)

OUTPUT:

  • The composition of nps whose descents are listed in descents, assuming that nps is not None (otherwise, the last element of descents is removed from descents, and nps is set to be this last element plus 1).

EXAMPLES:

sage: [x-1 for x in Composition([1, 1, 3, 4, 3]).to_subset()]
[0, 1, 4, 8]
sage: Compositions().from_descents([1,0,4,8],12)
[1, 1, 3, 4, 3]
sage: Compositions().from_descents([1,0,4,8,11])
[1, 1, 3, 4, 3]

>>> from sage.all import *
>>> [x-Integer(1) for x in Composition([Integer(1), Integer(1), Integer(3), Integer(4), Integer(3)]).to_subset()]
[0, 1, 4, 8]
>>> Compositions().from_descents([Integer(1),Integer(0),Integer(4),Integer(8)],Integer(12))
[1, 1, 3, 4, 3]
>>> Compositions().from_descents([Integer(1),Integer(0),Integer(4),Integer(8),Integer(11)])
[1, 1, 3, 4, 3]
from_subset(S, n)[source]

The composition of \(n\) corresponding to the subset S of \(\{1, 2, \ldots, n-1\}\) under the bijection that maps the composition \((i_1, i_2, \ldots, i_k)\) of \(n\) to the subset \(\{i_1, i_1 + i_2, i_1 + i_2 + i_3, \ldots, i_1 + \cdots + i_{k-1}\}\) (see Composition.to_subset()).

INPUT:

  • S – an iterable, a subset of \(\{1, 2, \ldots, n-1\}\)

  • n – integer

EXAMPLES:

sage: Compositions().from_subset([2,1,5,9], 12)
[1, 1, 3, 4, 3]
sage: Compositions().from_subset({2,1,5,9}, 12)
[1, 1, 3, 4, 3]

sage: Compositions().from_subset([], 12)
[12]
sage: Compositions().from_subset([], 0)
[]

>>> from sage.all import *
>>> Compositions().from_subset([Integer(2),Integer(1),Integer(5),Integer(9)], Integer(12))
[1, 1, 3, 4, 3]
>>> Compositions().from_subset({Integer(2),Integer(1),Integer(5),Integer(9)}, Integer(12))
[1, 1, 3, 4, 3]

>>> Compositions().from_subset([], Integer(12))
[12]
>>> Compositions().from_subset([], Integer(0))
[]
class sage.combinat.composition.Compositions_all[source]

Bases: Compositions

Class of all compositions.

subset(size=None)[source]

Return the set of compositions of the given size.

EXAMPLES:

sage: C = Compositions()
sage: C.subset(4)
Compositions of 4
sage: C.subset(size=3)
Compositions of 3

>>> from sage.all import *
>>> C = Compositions()
>>> C.subset(Integer(4))
Compositions of 4
>>> C.subset(size=Integer(3))
Compositions of 3
zero()[source]

Return the zero of the additive monoid.

This is the empty composition.

EXAMPLES:

sage: C = Compositions()
sage: C.zero()
[]

>>> from sage.all import *
>>> C = Compositions()
>>> C.zero()
[]
class sage.combinat.composition.Compositions_constraints(*args, **kwds)[source]

Bases: IntegerListsLex

class sage.combinat.composition.Compositions_n(n)[source]

Bases: Compositions

Class of compositions of a fixed \(n\).

cardinality()[source]

Return the number of compositions of \(n\).

random_element()[source]

Return a random Composition with uniform probability.

This method generates a random binary word starting with a 1 and then uses the bijection between compositions and their code.

EXAMPLES:

sage: Compositions(5).random_element() # random
[2, 1, 1, 1]
sage: Compositions(0).random_element()
[]
sage: Compositions(1).random_element()
[1]

>>> from sage.all import *
>>> Compositions(Integer(5)).random_element() # random
[2, 1, 1, 1]
>>> Compositions(Integer(0)).random_element()
[]
>>> Compositions(Integer(1)).random_element()
[1]
sage.combinat.composition.composition_iterator_fast(n)[source]

Iterator over compositions of \(n\) yielded as lists.