Two-color partitions with evens in one color Thanks: First author partially supported by Simons Foundation Grant 633284 George E. Andrews and Mohamed El Bachraoui Abstract . We consider sequences counting integer partitions in two colors (red and blue) in which the even parts occur only in blue color. We focus on subsequences defined by constraints on the parity and color of the summands. We establish formulas for our sequences and deduce identities of integer partitions. 2000 Mathematics Subject Classification 11P81; 05A17; 11D09 1. Introduction We adopt the following standard notation from the theory of q -series [ 1 , 9 ] (a;q)_{0}=1,\ (a;q)_{n}=\prod_{j=0}^{n-1}(1-aq^{j}),\quad(a;q)_{\infty}=\prod_{j=0}^{\infty}(1-aq^{j}), (a_{1},\ldots,a_{k};q)_{n}=\prod_{j=1}^{k}(a_{j};q)_{n},\ (a_{1},\ldots,a_{k};q)_{\infty}=\prod_{j=1}^{k}(a_{j};q)_{\infty}, and {}_{2}\phi_{1}\biggl[\genfrac{.}{.}{0.0pt}{}{a,b}{c};q\mathchar 59\relax\mkern 6.0muz\biggr]:=\sum_{n\geq 0}\frac{(a,b;q)_{n}z^{n}}{(q,c;q)_{n}}. In this paper we consider sequences of integer partitions in two colors (red and blue) with conditions on the parity and color of their summands. We shall write \lambda_{b} (resp. \lambda_{r} ) to denote a part \lambda occurring in blue (resp. red) color assuming the following order \lambda_{b}\geq\lambda_{r} . For references on two-color partitions, see for instance [ 2 , 3 , 4 , 5 ] . All of our sequences come from the following set. Definition 1 . For any nonnegative integer n , let \mathcal{F}(n) be the set of two-color integer partitions n such that the even parts may occur only in blue color and let F(n)=|\mathcal{F}(n)| . Then we clearly have \sum_{n\geq 0}F(n)q^{n}=\frac{1}{(q;q^{2})_{\infty}^{2}(q^{2};q^{2})_{\infty}}=1+2q+4q^{2}+8q^{3}+14q^{4}+24q^{5}+\cdots. (1.1) Definition 2 . For a nonnegative integer n , let F_{0}(n) (resp. F_{1}(n) ) be the number of partitions in \mathcal{F}(n) in which the number of odd parts in red color is even (resp. odd). Furthermore, let F_{2}(n) (resp. F_{3}(n) ) be the number of partitions in \mathcal{F}(n) in which the number of even parts is even (resp. odd). Our first goal in this paper is to establish the following two results regarding F_{0}(n) and F_{1}(n) . Theorem 1 . There holds \sum_{n\geq 0}F_{0}(n)q^{n}=\frac{(q^{16},q^{6},q^{10};q^{16})_{\infty}}{(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}^{2}}. We get the following immediate consequence of Theorem 1 . Corollary 1 . For any nonnegative integer n , we have that F_{0}(n) equals the number of partitions of n in two colors (red and blue) wherein the odd parts and the parts \equiv 6,10,16\ (\mathrm{mod}\ 16) may appear only in the blue color. Theorem 2 . There holds \sum_{n\geq 0}F_{1}(n)q^{n}=q\frac{(q^{16},q^{2},q^{14};q^{16})_{\infty}}{(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}^{2}}. We get the following immediate corollary of Theorem 2 . Corollary 2 . For any nonnegative integer n , we have that F_{1}(n) equals the number of partitions of n-1 in two colors (red and blue) wherein the odd parts and the parts \equiv 2,14,16\ (\mathrm{mod}\ 16) may appear only in the blue color. For our next two theorems we need the following definitions. Let A be a positive integer, let a be a nonnegative integer, and let \pi be partition of n . Andrews and Newman recently in [ 6 ] defined the minimal excludant function \mex_{A,a}(\pi) to be the smallest integer \equiv a\ (\mathrm{mod}\ A) which is not part of \pi . Then they defined p_{A,a}(n) (resp. \overline{p}_{A,a}(n) ) as the number of partitions \pi of n , where \mex_{A,a}(\pi)\equiv a\ (\mathrm{mod}\ 2A)\quad\text{(resp. $\mex_{A,a}(\pi)\equiv A+a\ (\mathrm{mod}\ 2A)$)}. We now extend these concepts to the two-color partitions in the set \mathcal{F}(n) as follows. Assume that A is even and that the parts \equiv 0\ (\mathrm{mod}\ \frac{A}{2}) may occur only in the blue color. Let \mex_{A,a}(\pi,\text{blue}) be the smallest integer \equiv a\ (\mathrm{mod}\ A) which is not part of \pi . Accordingly, let p_{A,a}(n,\text{blue}) (resp. \overline{p}_{A,a}(n,\text{blue}) ) count the number of partitions \pi in \mathcal{F}(n) , where \mex_{A,a}(\pi,\text{blue})\equiv a\ (\mathrm{mod}\ 2A)\quad\text{(resp. $\mex_{A,a}(\pi,\text{blue})\equiv A+a\ (\mathrm{mod}\ 2A)$)}. Our second goal is to prove the following partition identities. Theorem 3 . For any nonnegative integer n , we have that F_{2}(n)=p_{4,2}(n,\text{blue}) . Theorem 4 . For any nonnegative integer n , we have that F_{3}(n)=\overline{p}_{4,2}(n,\text{blue}) . For example, for n=4 , we have F(n)=14 , enumerating 4_{b},3_{b}+1_{b},3_{b}+1_{r},3_{r}+1_{b},3_{r}+1_{r},2_{b}+2_{b},2_{b}+1_{b}+1_{b},2_{b}+1_{b}+1_{r},2_{b}+1_{r}+1_{r} 1_{b}+1_{b}+1_{b}+1_{b},1_{b}+1_{b}+1_{b}+1_{r},1_{b}+1_{b}+1_{r}+1_{r},1_{b}+1_{r}+1_{r}+1_{r},1_{r}+1_{r}+1_{r}+1_{r}. Then it is easy to check that F_{2}(4)=10 and F_{3}(4)=4 . Note that if n=4 , then the part 2 is the only possible missing part which is \equiv 2\ (\mathrm{mod}\ 4) . Then p_{4,2}(4,\text{blue})=10 , counting 4_{b},3_{b}+1_{b},3_{b}+1_{r},3_{r}+1_{b},3_{r}+1_{r},1_{b}+1_{b}+1_{b}+1_{b},1_{b}+1_{b}+1_{b}+1_{r},1_{b}+1_{b}+1_{r}+1_{r}, 1_{b}+1_{r}+1_{r}+1_{r},1_{r}+1_{r}+1_{r}+1_{r}. The confirms that F_{2}(4)=p_{4,2}(4,\text{blue}) . Furthermore, we have \overline{p}_{4,2}(4,\text{blue})=4 counting 2_{b}+2_{b},2_{b}+1_{b}+1_{b},2_{b}+1_{b}+1_{r},2_{b}+1_{r}+1_{r}. This shows that F_{3}(4)=\overline{p}_{4,2}(4,\text{blue}) . Finally, it is worth noting that F(n) arises naturally in the theory of integer partitions. Indeed, by Euler’s formula [ 1 ] (-q;q)_{\infty}=\frac{1}{(q;q^{2})_{\infty}}, (1.2) we get \displaystyle\sum_{n\geq 0}F(n)q^{n} \displaystyle=\frac{1}{(q;q^{2})_{\infty}^{2}(q^{2};q^{2})_{\infty}}=\frac{(-q;q)_{\infty}}{(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}}=\frac{(-q;q)_{\infty}}{(q;q)_{\infty}} \displaystyle=\sum_{n\geq 0}\overline{p}(n)q^{n}, (1.3) where \overline{p}(n) is the number of overpartitions of n . Here, by an overpartition of n we mean a partition of n where the first occurrence of a part may be overlined, see [ 8 ] . Our third goal in this note is to express some of our sequences and their subsequences in terms of overpartitions. This rest of the paper is organized as follows. In Section 2 , we establish more connections between our sequences and integer overpartitions. In Section 3 we provide the proof of Theorems 1 - 2 . In Section 4 we give the proof of Theorems 3 - 4 . Finally, Section 5 is devoted to some comments including new q -series identities and open problems which are related to our work in this paper. 2. Connections with overpartitions We will assume that \overline{p}(k)=0 if k is not a nonnegative integer. we start with the following formulas for F_{0}(n) and F_{1}(n) in terms of \overline{p}(n) . Theorem 5 . For any nonnegative integer n there holds \displaystyle F_{0}(n) \displaystyle=\frac{\overline{p}(n)+\overline{p}\big(\frac{n}{2}\big)}{2}, (2.1) \displaystyle F_{1}(n) \displaystyle=\frac{\overline{p}(n)-\overline{p}\big(\frac{n}{2}\big)}{2},. (2.2) Proof. It easily seen that \sum_{n\geq 0}\big(F_{0}(n)-F_{1}(n)\big)q^{n}=\frac{1}{(-q;q^{2})_{\infty}(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}}. Then by ( 1.2 ) and ( 1.3 ), we obtain \displaystyle\sum_{n\geq 0}\big(F_{0}(n)-F_{1}(n)\big)q^{n} \displaystyle=\frac{1}{(-q;q^{2})_{\infty}(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}} \displaystyle=\frac{(-q;q)_{\infty}}{(-q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}} \displaystyle=\frac{(-q^{2};q^{2})_{\infty}}{(q^{2};q^{2})_{\infty}} \displaystyle=\sum_{n\geq 0}\overline{p}(n)q^{2n}. (2.3) Furthermore, we clearly have \sum_{n\geq 0}\big(F_{0}(n)+F_{1}(n)\big)q^{n}=\sum_{n\geq 0}F(n)q^{n}, and thus by ( 1.3 ), \sum_{n\geq 0}\big(F_{0}(n)+F_{1}(n)\big)q^{n}=\sum_{n\geq 0}\overline{p}(n)q^{n}. (2.4) Now combine ( 3.1 ) with ( 3.2 ) to achieve the desired formulas. ∎ We now introduce a subset of \mathcal{F}(n) . Definition 3 . For any nonnegative integer n , let \mathcal{H}(n) be the subset of \mathcal{F}(n) wherein the parts of the same color do not repeat and let H(n)=|\mathcal{H}(n)| . Then it is easy to see that \sum_{n\geq 0}H(n)q^{n}=(-q;q^{2})_{\infty}^{2}(-q^{2};q^{2})_{\infty}=1+2q+2q^{2}+4q^{3}+6q^{4}+8q^{5}+\cdots. (2.5) Now by ( 1.2 ), we find \displaystyle\sum_{n\geq 0}H(n)q^{n} \displaystyle=(-q;q^{2})_{\infty}^{2}(-q^{2};q^{2})_{\infty} \displaystyle=\frac{(-q;q^{2})_{\infty}}{(q;q^{2})_{\infty}} \displaystyle=\sum_{n\geq 0}\overline{p}_{o}(n)q^{n}, (2.6) where \overline{p}_{o}(n) is the number of overpartitions of n into odd parts, [ 7 , 10 ] . For example, for n=4 we have H(4)=6 counting 4_{b},3_{b}+1_{b},3_{b}+1_{r},3_{r}+1_{b},3_{r}+1_{r},2_{b}+1_{b}+1_{r} and \overline{p}_{o}(4)=6 , counting \bar{3}+\bar{1},\bar{3}+1,3+\bar{1},3+1,\bar{1}+1+1+1,1+1+1+1. Definition 4 . For a nonnegative integer n , let H_{0}(n) (resp. H_{1}(n) ) be the number of partitions counted by H(n) in which the number of even parts is even (resp. odd). Furthermore, let H_{2}(n) (resp. H_{3}(n) ) be the number of partitions counted by H(n) in which the number of parts is even (resp. odd). Theorem 6 . For any nonnegative integer n , there holds \begin{split}(a)\ &H_{0}(n)=\begin{cases}\frac{\overline{p}_{o}(n)}{2}+1&\text{if $n$ is a square,}\\ \frac{\overline{p}_{o}(n)}{2}&\text{otherwise,}\end{cases}\\ (b)\ &H_{1}(n)=\begin{cases}\frac{\overline{p}_{o}(n)}{2}-1&\text{if $n$ is a square,}\\ \frac{\overline{p}_{o}(n)}{2}&\text{otherwise,}\end{cases}\\ (c)\ &H_{2}(n)=\begin{cases}\frac{\overline{p}_{o}(n)}{2}+(-1)^{n}&\text{if $n$ is a square,}\\ \frac{\overline{p}_{o}(n)}{2}&\text{otherwise,}\end{cases}\\ (d)\ &H_{3}(n)=\begin{cases}\frac{\overline{p}_{o}(n)}{2}-(1)^{n}&\text{if $n$ is a square,}\\ \frac{\overline{p}_{o}(n)}{2}&\text{otherwise.}\end{cases}\end{split} Proof. (a, b) It clear that \sum_{n\geq 0}\big(H_{0}(n)-H_{1}(n)\big)q^{n}=(q^{2};q^{2})_{\infty}(-q;q^{2})_{\infty}^{2}, which by ( 4.1 ) applied with q\to q^{2} and x=-q^{-1} yields \sum_{n\geq 0}\big(H_{0}(n)-H_{1}(n)\big)q^{n}=1+2\sum_{n=1}^{\infty}q^{n^{2}}. (2.7) Furthermore, by ( 2.5 ) and ( 2.6 ) \displaystyle\sum_{n\geq 0}\big(H_{0}(n)+H_{1}(n)\big)q^{n} \displaystyle=(-q;q^{2})_{\infty}^{2}(-q^{2};q^{2})_{\infty} \displaystyle=\sum_{n\geq 0}H(n)q^{n} \displaystyle=\sum_{n=0}^{\infty}\overline{p}_{0}(n)q^{n}. (2.8) Then the desired identities for H_{0}(n) and H_{1}(n) follow by combining ( 2.7 ) and ( 2.8 ). (c, d) Similarly, with the help of ( 4.1 ) applied to q\to q^{2} and x=-q^{-1} , we get \sum_{n\geq 0}\big(H_{2}(n)-H_{3}(n)\big)q^{n}=(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}=1+2\sum_{n=1}^{\infty}(-1)^{n}q^{n^{2}}. (2.9) In addition, with the help of ( 2.6 ), \sum_{n\geq 0}\big(H_{2}(n)+H_{3}(n)\big)q^{n}=\sum_{n\geq 0}H(n)q^{n}=\sum_{n=0}^{\infty}\overline{p}_{o}(n)q^{n}. (2.10) Now use ( 2.9 ) and ( 2.10 ) to derive the desired identities for H_{2}(n) and H_{3}(n) . ∎ 3. Proof of Theorems 1 - 2 Proof of Theorem 1 . It is easily seen that \displaystyle\sum_{n\geq 0}\big(F_{0}(n)-F_{1}(n)\big)q^{n} \displaystyle=\frac{1}{(-q;q^{2})_{\infty}(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}}, \displaystyle=\frac{1}{(-q;q^{2})_{\infty}(q;q)_{\infty}}, (3.1) \displaystyle\sum_{n\geq 0}\big(F_{0}(n)+F_{1}(n)\big)q^{n} \displaystyle=\frac{1}{(q;q^{2})_{\infty}(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}}, \displaystyle=\frac{1}{(q;q^{2})_{\infty}(q;q)_{\infty}}. (3.2) Then upon adding ( 3.1 ) and ( 3.2 ), we find \displaystyle\sum_{n\geq 0}F_{0}(n)q^{n} \displaystyle=\frac{1}{2}\frac{1}{(q;q)_{\infty}}\Big(\frac{1}{(q;q^{2})_{\infty}}+\frac{1}{(-q;q^{2})_{\infty}}\Big) \displaystyle=\frac{1}{2}\frac{1}{(q;q)_{\infty}(q^{2};q^{4})_{\infty}}\big((-q;q^{2})_{\infty}+(q;q^{2})_{\infty}\big) \displaystyle=\frac{1}{2}\frac{1}{(q;q)_{\infty}(q^{2};q^{2})_{\infty}}\big((-q,-q^{3},q^{4};q^{4})_{\infty}+(q,q^{3},q^{4};q^{4})_{\infty}\big) \displaystyle=\frac{1}{2}\frac{1}{(q;q)_{\infty}(q^{2};q^{2})_{\infty}}\sum_{n=-\infty}^{\infty}q^{2n^{2}+n}\big(1+(-1)^{n}\big) (3.3) \displaystyle=\frac{1}{(q;q)_{\infty}(q^{2};q^{2})_{\infty}}\sum_{n=-\infty}^{\infty}q^{8n^{2}+2n} \displaystyle=\frac{(q^{16},q^{6},q^{10};q^{16})_{\infty}}{(q;q)_{\infty}(q^{2};q^{2})_{\infty}} (3.4) \displaystyle=\frac{(q^{16},q^{6},q^{10};q^{16})_{\infty}}{(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}^{2}}, where in ( 3.3 ) we applied ( 4.1 ) with q\to q^{4} and x=q^{-1} and in ( 3.4 ) we applied ( 4.1 ) with q\to q^{16} and x=q^{-6} . This completes the proof. Proof of Theorem 1 . By subtracting ( 3.1 ) from ( 3.2 ), we similarly get \displaystyle\sum_{n\geq 0}F_{1}(n)q^{n} \displaystyle=\frac{1}{2}\frac{1}{(q;q)_{\infty}}\Big(\frac{1}{(q;q^{2})_{\infty}}-\frac{1}{(-q;q^{2})_{\infty}}\Big) \displaystyle=\frac{1}{2}\frac{1}{(q;q)_{\infty}(q^{2};q^{2})_{\infty}}\sum_{n=-\infty}^{\infty}q^{2n^{n}+n}\big(1-(-1)^{n}\big) \displaystyle=\frac{1}{(q;q)_{\infty}(q^{2};q^{2})_{\infty}}\sum_{n=-\infty}^{\infty}q^{8n^{2}+6n+1} \displaystyle=q\frac{(q^{16},q^{2},q^{14};q^{16})_{\infty}}{(q;q^{2})_{\infty}(q^{2};q^{2})_{\infty}^{2}}. This completes the proof. 4. Proof of Theorems 3 - 4 We need Jacobi’s triple product identity [ 9 , p. 12] (q,-xq,-1/x;q)_{\infty}=\sum_{n=-\infty}^{\infty}x^{n}q^{\frac{n(n+1)}{2}}. (4.1) Proof of Theorem 3 . It clear that \displaystyle\sum_{n\geq 0}\big(F_{2}(n)-F_{3}(n)\big)q^{n} \displaystyle=\frac{1}{(-q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}, (4.2) \displaystyle\sum_{n\geq 0}\big(F_{2}(n)+F_{3}(n)\big)q^{n} \displaystyle=\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}. (4.3) Then by adding ( 4.2 ) and ( 4.3 ) we get \displaystyle\sum_{n\geq 0}F_{2}(n)q^{n} \displaystyle=\frac{1}{2}\Big(\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}+\frac{1}{(-q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\Big) \displaystyle=\frac{1}{2(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\Big(1+\frac{(q^{2};q^{2})_{\infty}}{(-q^{2};q^{2})_{\infty}}\Big) \displaystyle=\frac{1}{2(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\big(1+(q^{2};q^{2})_{\infty}(q^{2};q^{4})_{\infty}\big) (4.4) \displaystyle=\frac{1}{2(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\big(1+(q^{4},q^{2},q^{2};q^{4})_{\infty}\big) \displaystyle=\frac{1}{2(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\Big(1+\sum_{n=-\infty}^{\infty}(-1)^{n}q^{2n^{2}}\Big) (4.5) \displaystyle=\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 0}(-1)^{n}q^{2n^{2}} \displaystyle=\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 0}q^{8n^{2}}(1-q^{8n+2}) \displaystyle=\frac{1}{(q^{4};q^{4})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 0}\frac{q^{8n^{2}}(1-q^{8n+2})}{(q^{2};q^{4})_{\infty}} \displaystyle=\frac{1}{(q^{4};q^{4})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 0}\frac{q^{2+6+10+\ldots+(8n-2)}}{\displaystyle{\prod_{\begin{subarray}{c}j=0\\ j\neq 2n\end{subarray}}^{\infty}}(1-q^{4j+2})}, where in step ( 4.4 ) we used ( 1.2 ) and in step ( 4.5 ) we applied ( 4.1 ) with q\to q^{4} and x=-q^{-2} . This completes the proof. Proof of Theorem 4 . Similarly, by subtracting ( 4.2 ) from ( 4.3 ), we obtain \displaystyle\sum_{n\geq 0}F_{3}(n)q^{n} \displaystyle=\frac{1}{2}\Big(\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}-\frac{1}{(-q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\Big) \displaystyle=\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 1}(-1)^{n}q^{2n^{2}} \displaystyle=\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 0}(-1)^{n+1}q^{2(n+1)^{2}} \displaystyle=\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 0}q^{8n^{2}+8n+2}(1-q^{8n+6}) \displaystyle=\frac{1}{(q^{4};q^{4})_{\infty}(q;q^{2})_{\infty}^{2}}\sum_{n\geq 0}\frac{q^{2+6+10+\ldots+(8n+2)}}{\displaystyle{\prod_{\begin{subarray}{c}j=0\\ j\neq 2n+1\end{subarray}}^{\infty}}(1-q^{4j+2})}. This completes the proof. 5. Concluding remarks 1. Our proofs for Corollaries 1 - 2 and for Theorems 3 - 4 are analytic relying of q -series manipulations. Open Problem 1 . Give bijective proofs for Corollaries 1 - 2 and for Theorems 3 - 4 . 2. Recall that F(n) is the number of two-color integer partitions n such that the even parts may occur only in blue color. Now, by separating the cases according to whether the greatest part is even–blue, odd–blue, or odd–red, and keeping in mind the assumed order \lambda_{b}>\lambda_{r} , we find that \displaystyle\sum_{n\geq 0}F(n)q^{n} \displaystyle=\sum_{n\geq 0}\frac{q^{2n}}{(q^{2};q^{2})_{n}(q;q^{2})_{n}^{2}}\quad(\text{greatest part even}) \displaystyle+\sum_{n\geq 0}\frac{q^{2n+1}}{(q^{2};q^{2})_{n}(q;q^{2})_{n+1}^{2}}\quad(\text{greatest part odd-blue}) \displaystyle+\sum_{n\geq 0}\frac{q^{2n+1}}{(q^{2};q^{2})_{n}(q;q^{2})_{n+1}(q;q^{2})_{n}}\quad(\text{greatest part odd-red}). (5.1) On the other hand, if we consider the smallest part instead of the greatest part, we get \displaystyle\sum_{n\geq 0}F(n)q^{n} \displaystyle=\sum_{n\geq 0}\frac{q^{2n+2}}{(q^{2n+2},q^{2n+3},q^{2n+3};q^{2})_{\infty}}\quad(\text{smallest part even}) \displaystyle+\sum_{n\geq 0}\frac{q^{2n+1}}{(q^{2n+2},q^{2n+1},q^{2n+3};q^{2})_{\infty}}\quad(\text{smallest part odd-blue}) \displaystyle+\sum_{n\geq 0}\frac{q^{2n+1}}{(q^{2n+2},q^{2n+1},q^{2n+1};q^{2})_{\infty}}\quad(\text{smallest part odd-red}). (5.2) Thus, by a combination of ( 5.1 )-( 5.2 ) and ( 1.1 ), we get the following identities of q -series. \displaystyle\sum_{n\geq 0}\frac{q^{2n}}{(q^{2};q^{2})_{n}(q;q^{2})_{n}^{2}}\Big(1+\frac{q}{1-q^{2n+1}}+\frac{q}{(1-q^{2n+1})^{2}}\Big) \displaystyle=1+\sum_{n\geq 0}\frac{q^{2n+1}}{(q^{2n+2};q^{2})_{\infty}(q^{2n+3};q^{2})_{\infty}^{2}}\Big(q+\frac{1}{1-q^{2n+1}}+\frac{1}{(1-q^{2n+1})^{2}}\Big) \displaystyle=\frac{1}{(q^{2};q^{2})_{\infty}(q;q^{2})_{\infty}^{2}}. (5.3) Similarly, recall that H(n) is the number of two-color integer partitions n such that the even parts may occur only in blue color and the parts of the same color are not repeated. Now, separating the cases according to whether the greatest part is even–blue, odd–blue, or odd–red, we find that \displaystyle\sum_{n\geq 1}H(n)q^{n} \displaystyle=1+\sum_{n\geq 0}q^{2n+2}(-q^{2};q^{2})_{n}(-q;q^{2})_{n+1}^{2}\quad(\text{greatest part even}) \displaystyle+\sum_{n\geq 0}q^{2n+1}(-q^{2};q^{2})_{n}(-q;q^{2})_{n}^{2}\quad(\text{greatest part odd-red}) \displaystyle+\sum_{n\geq 0}q^{2n+1}(-q^{2};q^{2})_{n}(-q;q^{2})_{n+1}(-q;q^{2})_{n}\quad(\text{greatest part odd-blue}). (5.4) Taking into account the smallest part instead of the greatest part, we deduce \displaystyle\sum_{n\geq 1}H(n)q^{n} \displaystyle=1+\sum_{n\geq 0}q^{2n+2}(-q^{2n+4},-q^{2n+3},-q^{2n+3};q^{2})_{\infty}\quad(\text{smallest part even}) \displaystyle+\sum_{n\geq 0}q^{2n+1}(-q^{2n+2},-q^{2n+1},-q^{2n+3};q^{2})_{\infty}\quad(\text{smallest part odd-red}) \displaystyle+\sum_{n\geq 0}q^{2n+1}(-q^{2n+2},-q^{2n+3},-q^{2n+3};q^{2})_{\infty}\quad(\text{smallest part odd-blue}). (5.5) Thus, combining ( 5.4 )-( 5.5 ) with ( 2.5 ), we derive the following identities of q -series \displaystyle\sum_{n\geq 0}q^{2n+1}(-q^{2};q^{2})_{n}(-q;q^{2})_{n+1}^{2}\Big(q+\frac{1}{1+q^{2n+1}}+\frac{1}{(1+q^{2n+1})^{2}}\Big) \displaystyle=\sum_{n\geq 0}q^{2n+1}(-q^{2n+1},-q^{2n+2},-q^{2n+3};q^{2})_{\infty}\Big(1+\frac{1}{1+q^{2n+1}}+\frac{q}{(1+q^{2n+1})(1+q^{2n+1})}\Big) \displaystyle=(-q^{2};q^{2})_{\infty}(-q;q^{2})_{\infty}^{2}-1. (5.6) Open Problem 2 . Give analytic proofs for the formulas listed in ( 5.3 ) and ( 5.6 ). Data Availability Statement. Not applicable. Declarations. The authors state that there is no conflict of interest.