WorksMathematics
Excedances and Deficiencies of Integer Partitions
Abstract
The notion of a fixed point, applied by Blecher and Knopfmacher to integer partitions, can be considered as a permutation statistic. Here we apply two related permutation statistics, excedance and deficiency, to integer partitions. Using analytic and combinatorial techniques, we establish several generating functions and make connections to partition concepts including Dyson's crank statistic, overpartitions, and the Rogers–Ramanujan partitions.
Introduction and background
A partition of a positive integer is a weakly decreasing sequence of positive integers whose sum is . The positive integers in the sequence are called parts. Let be the set of partitions of and write . By convention, .
The Durfee square of a partition has dimension where is the greatest index for which . This statistic gives a combinatorial understanding of Jacobi’s result
where we use standard -series notation; see, for example, Chapter 10 of Andrews, Askey, and Roy [1]. As shown in Figure 1, a partition can be decomposed into its Durfee square, the conjugate of a partition with largest part at most to the right of the Durfee square, and a partition with largest part at most below the Durfee square.
Figure 1. The Durfee square decomposition of a partition.
We will also use the series studied by Heine, in particular the identity [1 Equation 10.10.1]:
The crank of a partition, named by Dyson in 1944 but not found until 1988 work of Andrews and Garvan [3], is arguably the most important partition statistic. We will not need its definition, but will use the following generating function for the number of partitions of with positive crank. This was established in both Andrews–Newman [4 Theorem 1] and Hopkins–Sellers [8 proof of Theorem 1], 2020 works that independently established a connection between the crank and another partition statistic called the mex.
Proposition 1. A generating function for partitions with non-negative crank is
In 2022, Blecher and Knopfmacher applied the pervasive mathematical notion of fixed points to partitions as follows [5].
Definition 2. A partition has a fixed point if for some index .
Note that, due to the ordering convention, a partition has at most one fixed point. Also, if , then is the size of the Durfee square.
In resolving a conjecture of Blecher and Knopfmacher, Hopkins and Sellers established the following connections between fixed points and the crank [9].
Proposition 3. Given , the number of partitions of with a fixed point equals the number of partitions of with positive crank.
To refine the notion of fixed point, let be the number of partitions of that have a fixed point , i.e., . Hopkins established the following recurrence [7].
Proposition 4. For and ,
Considering fixed points in the context of permutations, partitions without a fixed point correspond to derangements. This setting invites the application of other permutation statistics to integer partitions. In this work, we consider two related permutation statistics on partitions.
Definition 5. An excedance of a partition is an index such that while a deficiency is an index such that . For , let be the number of partitions of with excedances, similarly for deficiencies.
The terminology dates at least to Reilly 1979 [11]. Note that every part of a partition is an excedance, fixed point, or deficiency. Further, any excedances are the initial parts of a partition, followed by a possible fixed point, with any deficiencies comprising the subsequent parts.
The next two sections give results on excedances and deficiencies, respectively, using both generating function and combinatorial techniques.
Excedances of integer partitions
The array of Figure 2 gives the number of partitions of with excedances for small values of and .

| 0 | 1 | 2 | 3 | 4 | |
| 0 | 1 | ||||
| 1 | 1 | ||||
| 2 | 1 | 1 | |||
| 3 | 1 | 2 | |||
| 4 | 1 | 4 | |||
| 5 | 1 | 6 | |||
| 6 | 1 | 9 | 1 | ||
| 7 | 1 | 12 | 2 | ||
| 8 | 1 | 16 | 5 | ||
| 9 | 1 | 20 | 9 | ||
| 10 | 1 | 25 | 16 | ||
| 11 | 1 | 30 | 25 | ||
| 12 | 1 | 36 | 39 | 1 | |
| 13 | 1 | 42 | 56 | 2 | |
| 14 | 1 | 49 | 80 | 5 | |
| 15 | 1 | 56 | 109 | 10 | |
| 16 | 1 | 64 | 147 | 19 | |
| 17 | 1 | 72 | 192 | 32 | |
| 18 | 1 | 81 | 249 | 54 | |
| 19 | 1 | 90 | 315 | 84 | |
| 20 | 1 | 100 | 396 | 129 | 1 |
Figure 2. Array of values for and .
Theorem 6. A bivariate generating function for the number of partitions of with excedances is
Proof. A partition with Durfee square size has excedances if , that is, referencing Figure 1, if is not a part of the subpartition to the right of the Durfee square. Otherwise, has excedances and is a part of (this case includes the empty partition of 0). A generating function for is then
which simplifies to the desired expression (3).
Working from each term of (5),
where we used (2) and, similarly,
which together give (4). □
The results of Theorem 6 allow us to explain many patterns in the array of Table 2.
| diag(Exc, 8) | |||
| Exc(8, 0) | (1, 1, 1, 1, 1, 1, 1, 1) | (1, 1, 1, 1, 1, 1, 1, 1, 1) | |
| Exc(7, 1) | (7) | (7, 2) | |
| (6, 1) | (6, 2, 1) | ||
| (5, 2) | (5, 2, 2) | ||
| (5, 1, 1) | (5, 2, 1, 1) | ||
| (4, 2, 1) | (4, 2, 2, 1) | ||
| (4, 1, 1, 1) | (4, 2, 1, 1, 1) | ||
| (3, 2, 2) | (3, 2, 2, 2) | ||
| (3, 2, 1, 1) | (3, 2, 2, 1, 1) | ||
| (3, 1, 1, 1, 1) | (3, 2, 1, 1, 1, 1) | ||
| (2, 2, 2, 1) | (2, 2, 2, 2, 1) | ||
| (2, 2, 1, 1, 1) | (2, 2, 2, 1, 1, 1) | ||
| (2, 1, 1, 1, 1, 1) | (2, 2, 1, 1, 1, 1) | ||
| Exc(6, 2) | (3, 3) | (3, 3, 3) |
Table 2. The example of the bijection in the second proof of Corollary 10.
Clearly the row sums of the array match , the number of partitions of . Analytically, letting in (3) gives Jacobi’s (1).
Corollary 7. A generating function for the number of partitions of with an even number of excedances is
and an analogue for odd number of excedances is
giving an expression for the difference
Proof. The excedance parity results follow from standard techniques applied to (4) and setting . The difference expression comes most easily from setting in (4). □
Table 1 provides initial values for these counts. Computations suggest that the difference remains positive for . At the time of writing, these sequences were not contained in the On-Line Encyclopedia of Integer Sequences [10].
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | |
| even | 1 | 1 | 1 | 1 | 1 | 2 | 3 | 6 | 10 | 17 | 26 | 40 | 57 | 81 | 110 |
| odd | 0 | 1 | 2 | 4 | 6 | 9 | 12 | 16 | 20 | 25 | 30 | 37 | 44 | 54 | 66 |
| diff. | 1 | 0 | 3 | 13 | 27 | 44 |
Table 1. The number of partitions of with an even number of excedances, an odd number, and the difference, through .
The weighted row sum gives the total number of excedances in the partitions of .
Corollary 8. A generating function for the total number of excedances in the partitions of is
Proof. The result follows from determining the partial derivative of (4) with respect to and evaluating that expression at . □
Values for for small are given in Table 5 in the next section.
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | |
| 0 | 1 | 2 | 4 | 6 | 11 | 16 | 26 | 38 | 57 | 80 | 117 | 160 | 224 | 304 | |
| 0 | 1 | 3 | 6 | 11 | 19 | 31 | 50 | 76 | 116 | 169 | 247 | 349 | 494 | 682 |
Table 5. The total number of excedances and deficiencies in for small .
Both the diagonal sums
and antidiagonal sums
(finite due to the shape of the array) have combinatorial interpretations.
Corollary 9. For , the diagonal sum equals the number of partitions of with positive crank.
Proof. Letting in (4) gives the diagonal sums of the array. The resulting expression reduces to
which is the complement of the shifted generating function of Proposition 1, i.e., the expression here is a generating function for the number of partitions of with positive crank, using crank symmetry [3]. □
Corollary 10. The antidiagonal sum equals .
Proof. Letting in (4) gives the antidiagonal sums of the array. The resulting term in the numerator is nonzero only for so that the expression reduces to
a generating function for . □
We also give combinatorial proofs for Corollaries 9 and 10. Let be the set of partitions of with excedances. Write for the set of partitions with a fixed point and for the set of partitions in with a fixed point [7].
Theorem 11. There is a bijection
Proof. Given , the excedances are necessarily the first parts. Note that and (else there are at least excedances). Assign
a partition of with fixed point so that . The reverse map is clear: remove the fixed point from to get . □
Examples of the bijection of Theorem 11 are given in Table 2.
Combinatorial proof of Corollary 9. By Theorem 11, the diagonal sum
the number of partitions of with a fixed point. By Proposition 3, this matches the number of partitions of with positive crank. □
Table 2 gives an example of the bijection in this combinatorial proof of Corollary 9 which uses three applications of the bijection of Theorem 11.
Theorem 11 also translates Proposition 4 into a recurrence for the array.
Corollary 12. For and ,
For the next combinatorial proof, note that either has Durfee square size with no fixed point, i.e., , or Durfee square size with fixed point .
Combinatorial proof of Corollary 10. We establish a bijection
For with Durfee square size , we know and so that reducing each excedance by one gives a valid partition
For with Durfee square size , we know and removing that fixed point leaves
Applying the map to , the first case gives partitions in and the second case gives partitions in .
For the reverse map, given with Durfee square size , increase the first parts by one to make with excedances, Durfee square size , and no fixed point. Given , find the greatest such that and add a part to make a partition in with fixed point and excedances. The bijection establishes the identity . □
Table 3 gives an example of the bijection in this combinatorial proof of Corollary 10.
Table 3. The example of the bijection in the second proof of Corollary 10.
Note that, by adding an overlined part 1 to the partitions of , Corollary 10 can be cast as a bijection to the restricted overpartitions of [6] where only a part 1 may be overlined.
Deficiencies of integer partitions
For permutations, excedances and deficiencies are essentially equivalent, since each excedance of a permutation is a deficiency of the inverse permutation and vice versa. That means that the number of excedances over all permutation of equals the number of deficiencies over . The situation for partitions is very different. As explained in the proof of Theorem 6, a partition of with Durfee square size has or excedances while the number of deficiencies, the number of parts in below the Durfee square as in Figure 1, could be as high as . See Corollary 16 and Table 5 for more details.
The array of Figure 3 gives the number of partitions of with deficiencies for small values of and .
Theorem 13. A bivariate generating function for the number of partitions of with deficiencies is
Proof. The number of deficiencies of a partition is exactly the number of parts of the subpartition underneath the Durfee square, which establishes (6).
| 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | |
| 1 | 1 | ||||||||||||||
| 2 | 1 | 1 | |||||||||||||
| 3 | 1 | 1 | 1 | ||||||||||||
| 4 | 2 | 1 | 1 | 1 | |||||||||||
| 5 | 2 | 2 | 1 | 1 | 1 | ||||||||||
| 6 | 3 | 3 | 2 | 1 | 1 | 1 | |||||||||
| 7 | 3 | 4 | 3 | 2 | 1 | 1 | 1 | ||||||||
| 8 | 4 | 5 | 5 | 3 | 2 | 1 | 1 | 1 | |||||||
| 9 | 5 | 6 | 6 | 5 | 3 | 2 | 1 | 1 | 1 | ||||||
| 10 | 6 | 8 | 8 | 7 | 5 | 3 | 2 | 1 | 1 | 1 | |||||
| 11 | 7 | 10 | 10 | 9 | 7 | 5 | 3 | 2 | 1 | 1 | 1 | ||||
| 12 | 9 | 13 | 13 | 12 | 10 | 7 | 5 | 3 | 2 | 1 | 1 | 1 | |||
| 13 | 10 | 16 | 17 | 15 | 13 | 10 | 7 | 5 | 3 | 2 | 1 | 1 | 1 | ||
| 14 | 12 | 20 | 22 | 20 | 17 | 14 | 10 | 7 | 5 | 3 | 2 | 1 | 1 | 1 | |
| 15 | 14 | 24 | 28 | 26 | 22 | 18 | 14 | 10 | 7 | 5 | 3 | 2 | 1 | 1 | 1 |
Table 3 (PDF p. 8). Array of values for .
Similar to the derivations of Theorem 6, we use (2) here:
as desired. □
It is interesting to contrast the rather complicated expressions (4) and (7) to
where each part is weighted by , not just the excedances or deficiencies.
Similar to the previous section, the results of Theorem 13 allow us to explain many patterns in the dfc array of Table 3.
Corollary 14. The sequence dfc equals the number of Rogers–Ramanujan partitions of , those restricted to parts congruent to 1 or 4 modulo 5.
Proof. Removing the term from the denominator in (6) leaves a generating function for the sequence, and the resulting
is a known generating function for the described Rogers–Ramanujan partitions [2 p. 52]. □
Corollary 15. A generating function for the number of partitions of with an even number of deficiencies is
and an analogue for odd number of deficiencies is
giving an expression for the difference
Proof. The deficiencies parity results follow from standard techniques applied to (7) and setting . The difference expression comes most easily from setting in (7). □
Table 4 provides initial values for the sequences described in Corollary 15. The even counts match [10 A046682] and odd counts match [10 A000701], while the difference matches the number of self-conjugate partitions [10 A000700].
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | |
| even | 1 | 1 | 2 | 3 | 4 | 6 | 8 | 12 | 16 | 22 | 29 | 40 | 52 | 69 | 90 |
| odd | 0 | 1 | 1 | 2 | 3 | 5 | 7 | 10 | 14 | 20 | 27 | 37 | 49 | 66 | 86 |
| diff. | 1 | 0 | 1 | 1 | 1 | 1 | 1 | 2 | 2 | 2 | 2 | 3 | 3 | 3 | 4 |
Table 4. The number of partitions of with an even number of deficiencies, an odd number, and the difference, through .
The weighted row sum gives the total number of deficiencies in the partitions of
Corollary 16. A generating function for the total number of deficiencies in the partitions of is
Proof. The result follows from determining the partial derivative of (7) with respect to and evaluating that expression at . □
Table 5 gives values of and for small , informing the discussion at the beginning of this section. The sequence is [10 A114089] (with this interpretation); the sequence is not in the OEIS at the time of writing.
Again, the diagonal sums
have a combinatorial interpretation (here the antidiagonal sums are infinite).
Corollary 17. The diagonal sum equals the number of partitions of with no part 2.
Proof. As in Corollary 9, letting in (7) gives the diagonal sums of the triangle. The resulting ratio
in the summation is nonzero only for and 1, leading to
where the latter is a generating function for partitions with no part 2. □
We will focus on more combinatorial aspects of partition deficiencies in future work.
References
- [1]G. E. Andrews, R. Askey, R. Roy (1999). Special Functions, Encyclopaedia of Mathematics and its Applications, vol. 71. Cambridge University Press.
- [2]G. E. Andrews, K. Eriksson (2004). Integer Partitions. Cambridge University Press.
- [3]G. E. Andrews, F. Garvan (1988). Dyson’s crank of a partition. Bull. Amer. Math. Soc. 18: 167–171.DOI
- [4]G. E. Andrews, D. Newman (2020). The minimal excludant in integer partitions. J. Integer Seq. 23: 20.2.3.
- [5]A. Blecher, A. Knopfmacher (2022). Fixed points and matching points in partitions, Ramanujan J. 58: 23–41.DOI
- [6]S. Corteel, J. Lovejoy (2004). Overpartitions. Trans. Amer. Math. Soc. 356: 1623–1635.
- [7]B. Hopkins (2024). Refining Blecher and Knopfmacher’s integer partition fixed points. Enumer. Combin. Appl. 4: S2R26.DOI
- [8]B. Hopkins, J. A. Sellers (2020). Turning the partition crank. Amer. Math. Monthly 127: 654–657.
- [9]B. Hopkins, J. A. Sellers (2024). On Blecher and Knopfmacher’s fixed points for integer partitions, Discrete Math. 347: 113839.DOI
- [10]OEIS Foundation Inc. (2026). The On-Line Encyclopedia of Integer Sequences, oeis.org.
- [11]J. W. Reilly (1979). A combinatorial solution of two related problems in sequence enumeration. J. Combin. Theory Ser. A 26: 304–307.DOI
