# Partial injections¶

Random partial injections and Stallings graphs

EXAMPLES:

sage: from slabbe import number_of_partial_injection
sage: number_of_partial_injection(10)
[1,
100,
4050,
86400,
1058400,
7620480,
31752000,
72576000,
81648000,
36288000,
3628800]


Random partial injections on [0, 1, ..., 6]:

sage: from slabbe import random_partial_injection
sage: random_partial_injection(7)
[None, None, 1, 3, None, 0, None]
sage: random_partial_injection(7)
[5, 1, 0, 3, None, 4, None]


Random Stallings graph on [0, 1, ..., 19] over 2 letters:

sage: from slabbe import random_cyclically_reduced_stallings_graph
sage: G,_,_ = random_cyclically_reduced_stallings_graph(20, 2)
sage: G
Looped multi-digraph on 20 vertices


Visualisation of the graph:

sage: from slabbe import TikzPicture
sage: tikz = TikzPicture.from_graph(G)
sage: path_to_file = tikz.pdf()    # not tested


REFERENCES:

• Bassino, Frédérique; Nicaud, Cyril; Weil, Pascal Random generation of finitely generated subgroups of a free group. Internat. J. Algebra Comput. 18 (2008), no. 2, 375–405.
slabbe.partial_injection.number_of_partial_injection(n, algorithm='binomial')

Return the number of partial injections on an set of $$n$$ elements defined on a subset of $$k$$ elements for each $$k$$ in $$0, 1, ..., n$$.

INPUT:

• n – integer
• algorithm – string (default: 'binomial'), 'binomial' or 'recursive'. When n>50, the binomial coefficient approach is faster (linear time vs quadratic time).

OUTPUT:

list

Note

The recursive code of this function was originally written by Vincent Delecroix (Nov 30, 2017) the day after a discussion with Pascal Weil and me at LaBRI.

EXAMPLES:

sage: from slabbe import number_of_partial_injection
sage: number_of_partial_injection(0)
[1]
sage: number_of_partial_injection(1)
[1, 1]
sage: number_of_partial_injection(2)
[1, 4, 2]
sage: number_of_partial_injection(3)
[1, 9, 18, 6]
sage: number_of_partial_injection(4)
[1, 16, 72, 96, 24]
sage: number_of_partial_injection(5)
[1, 25, 200, 600, 600, 120]
sage: number_of_partial_injection(6)
[1, 36, 450, 2400, 5400, 4320, 720]
sage: number_of_partial_injection(7)
[1, 49, 882, 7350, 29400, 52920, 35280, 5040]
sage: number_of_partial_injection(8)
[1, 64, 1568, 18816, 117600, 376320, 564480, 322560, 40320]


TESTS:

sage: number_of_partial_injection(8, algorithm='recursive')
[1, 64, 1568, 18816, 117600, 376320, 564480, 322560, 40320]


REFERENCE:

slabbe.partial_injection.random_cyclically_reduced_stallings_graph(n, r=2, verbose=False, merge=False)

Return a uniformly chosen Stallings graph of n vertices over r letters.

INPUT:

• n – integer, size of graph
• r – integer (default: 2), number of generators of the free group
• verbose – bool (default: False)

Note

The probability that G is connected is 1 - 2^r / n^(r-1) + o(1/n^(r-1)) which is approx. 1

OUTPUT:

digraph, integer, integer

EXAMPLES:

sage: from slabbe import random_cyclically_reduced_stallings_graph
sage: G,_,_ = random_cyclically_reduced_stallings_graph(20, 2)
sage: G
Looped multi-digraph on 20 vertices

sage: random_cyclically_reduced_stallings_graph(20, 5)[0]
Looped multi-digraph on 20 vertices


With verbose output:

sage: G = random_cyclically_reduced_stallings_graph(20, 2, verbose=True)   # random
rejecting because graph is not connected
rejecting because graph has a vertex of degree <=1
rejecting because graph has a vertex of degree <=1
rejecting because graph has a vertex of degree <=1


For displaying purposes, the following merges the multiedges automatically:

sage: G,_,_ = random_cyclically_reduced_stallings_graph(20, 2)
sage: from slabbe import TikzPicture
sage: tikz = TikzPicture.from_graph(G)
sage: _ = tikz.pdf(view=False)


AUTHORS:

• Sébastien Labbé and Pascal Weil, Dec 14, 2017, Sage Thursdays at LaBRI
slabbe.partial_injection.random_partial_injection(n)

Return a uniformly chosen random partial injection on 0, 1, …, n-1.

INPUT:

• n – integer

OUTPUT:

list

EXAMPLES:

sage: from slabbe import random_partial_injection
sage: random_partial_injection(10)
[3, 5, 2, None, 1, None, 0, 8, 7, 6]
sage: random_partial_injection(10)
[1, 7, 4, 8, 3, 5, 9, None, 6, None]
sage: random_partial_injection(10)
[5, 6, 8, None, 7, 4, 0, 9, None, None]


TODO:

Adapt the code once this is merged:

https://trac.sagemath.org/ticket/24416


AUTHORS:

• Sébastien Labbé and Vincent Delecroix, Nov 30, 2017, Sage Thursdays at LaBRI
slabbe.partial_injection.reject_statistics(n, r=2, sample_size=50, verbose=False)

Return return reject statistics when randomly chosing Stallings graph of n vertices over r letters.

INPUT:

• n – integer, size of graph
• r – integer (default: 2), number of generators of the free group
• n – integer (default: 50), size of sample
• verbose – bool (default: False)

OUTPUT:

histogram

EXAMPLES:

sage: from slabbe.partial_injection import reject_statistics
sage: h = reject_statistics(50, verbose=True)   # random
not connected: Counter({0: 48, 1: 2})
has degree 1: Counter({0: 27, 1: 18, 2: 3, 3: 1, 4: 1})
sage: h.save('h_50.png', title='size of graph=50') # not tested

sage: h = reject_statistics(100, verbose=True)  # random
not connected: Counter({0: 48, 1: 2})
has degree 1: Counter({0: 41, 1: 8, 2: 1})
sage: h.save('h_100.png', title='size of graph=100') # not tested

sage: h = reject_statistics(500, verbose=True)       # not tested (30s)
not connected: Counter({2: 5, 4: 5, 5: 5, 0: 4, 1: 4, 8: 4, 3: 3, 16:
3, 6: 2, 11: 2, 15: 2, 18: 2, 23: 2, 7: 1, 10: 1, 44: 1, 13: 1, 49:
1, 19: 1, 21: 1})
has degree 1: Counter({0: 14, 1: 9, 3: 8, 2: 5, 4: 3, 5: 3, 6: 2, 9: 2,
7: 1, 8: 1, 13: 1, 15: 1})
sage: h.save('h_500.png', title='size of graph=500') # not tested

sage: h = reject_statistics(1000, verbose=True)  # not tested (2min30s)
not connected: Counter({8: 4, 26: 3, 3: 2, 7: 2, 9: 2, 10: 2, 14: 2,
15: 2, 17: 2, 18: 2, 27: 2, 40: 2, 59: 2, 0: 1, 1: 1, 4: 1, 5: 1, 11:
1, 13: 1, 19: 1, 20: 1, 21: 1, 22: 1, 28: 1, 44: 1, 48: 1, 51: 1, 52:
1, 53: 1, 58: 1, 63: 1, 66: 1, 75: 1, 121: 1})
has degree 1: Counter({2: 9, 0: 7, 1: 6, 4: 6, 3: 4, 5: 4, 7: 4, 8: 2,
6: 1, 9: 1, 11: 1, 12: 1, 13: 1, 15: 1, 17: 1, 26: 1})
sage: h.save('h_1000.png', title='size of graph=1000') # not tested