Hall algebras associated to complexes of fixed size
Haicheng Zhang

TL;DR
This paper investigates the structure of Hall algebras formed from complexes of fixed size over projectives in a hereditary abelian category, establishing relations with cyclic complexes, derived Hall algebras, and providing an integration map.
Contribution
It introduces a detailed description of Hall algebras of fixed-size complexes, relating them to cyclic and derived Hall algebras, and constructs an explicit integration map.
Findings
Relation between Hall algebras of fixed size and cyclic complexes
Characterization of Hall algebra via generators and relations
Explicit integration map for 2-term complexes
Abstract
Let be a finitary hereditary abelian category with enough projectives. We study the Hall algebra of complexes of fixed size over projectives. Explicitly, we first give a relation between Hall algebras of complexes of fixed size and cyclic complexes. Secondly, we characterize the Hall algebra of complexes of fixed size by generators and relations, and relate it to the derived Hall algebra of . Finally, we give the integration map on the Hall algebra of -term complexes over projectives.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAlgebraic structures and combinatorial models · Advanced Topics in Algebra · Nonlinear Waves and Solitons
Hall algebras associated to complexes of fixed size
Haicheng Zhang
Institute of Mathematics, School of Mathematical Sciences, Nanjing Normal University, Nanjing 210023, P. R. China
Abstract.
Let be a finitary hereditary abelian category with enough projectives. We study the Hall algebra of complexes of fixed size over projectives. Explicitly, we first give a relation between Hall algebras of complexes of fixed size and cyclic complexes. Secondly, we characterize the Hall algebra of complexes of fixed size by generators and relations, and relate it to the derived Hall algebra of . Finally, we give the integration map on the Hall algebra of -term complexes over projectives.
Key words and phrases:
Bridgeland Hall algebra; derived Hall algebra; complex of fixed size.
2010 Mathematics Subject Classification:
17B37, 16G20, 17B20.
1. Introduction
The Hall algebra of a finite dimensional algebra over a finite field was introduced by Ringel [18] in 1990. Ringel [17, 18] proved that if is representation-finite and hereditary, the Ringel–Hall algebra of is isomorphic to the positive part of the corresponding quantized enveloping algebra. In order to give an intrinsic realization of the full quantized enveloping algebra via Hall algebra approach, one has managed to define the Hall algebra of a triangulated category satisfying some homological finiteness conditions (cf. [20], [21]). However, the root category of a finite dimensional algebra does not satisfy the homological finiteness conditions. In other word, the Hall algebra of a root category has not been defined.
In 2013, for each hereditary algebra , Bridgeland [4] introduced an algebra, called the Bridgeland Hall algebra of A, which is the Hall algebra of -cyclic complexes over projective -modules with some localization and reduction. He proved that the quantized enveloping algebra associated to A is embedded into the Bridgeland Hall algebra of . This provides a beautiful realization of the entire quantized enveloping algebra by Hall algebras.
Inspired by Bridgeland’s work, for each hereditary algebra and any nonnegative integer , Chen and Deng [8] applied Bridgeland’s construction to -cyclic complexes over projective -modules, and introduced the Bridgeland Hall algebra of -cyclic complexes of , whose algebra structure was characterized in [22].
Cluster algebras were introduced by Fomin and Zelevinsky in [10] and later the quantum cluster algebras were introduced by Berenstein and Zelevinsky in [3]. In [9], the author and his coauthors consider the Hall algebra of 2-term complexes over projective -modules of a finite acyclic quiver , and realize the quantum cluster algebra with principal coefficients as a quotient algebra of the achieved algebra with some localization and twist.
In this paper, let be a finitary abelian category with enough projectives and be a positive integer. The purpose of this paper is to generalize the construction of Hall algebra of 2-term complexes over projectives to -term complexes, and give a characterization on the algebra structure of the achieved algebra. In Section 2 we recall some homological properties of -cyclic complexes and -term complexes over projectives. We establish a relation between Hall algebras of -cyclic complexes and -term complexes in Section 3. From Section 4 on, we assume that is hereditary. We characterize indecomposable objects in the categories of -cyclic complexes and -term complexes over projectives in Section 4. In order to study the Hall algebra of -term complexes over projectives, in Section 5 we first give a characterization on the Hall algebra of bounded complexes over projectives, and relate it to the derived Hall algebra of . Section 6 is devoting to characterizing the Hall algebra of -term complexes over projectives by generators and relations. As an additional result, we give the integration map on the Hall algebra of -term complexes over projectives in Section 7.
Let us fix some notations used throughout the paper. Let be always a finite field with elements. Let be an (essentially small) finitary abelian -category with enough projectives, and be the subcategory consisting of projective objects. Given an exact category , we denote by , and the category of bounded complexes over , the bounded homotopy category, and the bounded derived category, respectively. The Grothendieck group of and the set of isomorphism classes of objects in are denoted by and , respectively. For any object we denote by the image of in . For a finite set , we denote by its cardinality. For an object in an additive category, we denote by the automorphism group of , and set .
2. Cyclic complexes and complexes of fixed size
In this section, we summarize some necessary homological properties of cyclic complexes and complexes of fixed size. We focus our attention on the complexes over projectives.
2.1. Cyclic complexes
For each positive integer , write . By definition, an -cyclic complex over consists of objects in and morphisms for satisfying . A morphism between two -cyclic complexes and is given by a family of morphisms satisfying for all . The category of -cyclic complexes over is denoted by . For notational simplicity, we write for the category of bounded complexes over , and set . The bounded complexes are called [math]-cyclic complexes. For each integer , we have a shift functor
[TABLE]
where is defined by
[TABLE]
For , let be the subcategory of , which is consisting of -cyclic complexes over . In the sense of component-wise exactness, is closed under extensions. For any morphism of projectives, if one defines by
[TABLE]
If one defines
[TABLE]
So each projective object determines an object in . By [8, Lemma 2.3], all indecomposable projective (injective) objects in are of the form for some indecomposable object and . That is, is a Frobenius exact category.
2.2. Complexes of fixed size
For each positive integer , we consider the category , it is the full subcategory of whose objects are the complexes with if That is,
[TABLE]
with for . Each object in is called an -term complex over . Let be the subcategory of , which is consisting of -term complexes over . Clearly, is an abelian category, and is closed under extensions. Actually, . For , the Auslander–Reiten theory and some homological properties of are studied in [2, 6, 7]. In fact, it is proved in [2] that is an exact category with enough projectives and injectives, and its global dimension is . Following [2], given an object , we consider the following objects in :
with if , and
with if , and ;
with if , and .
By [2, Corollary 3.9], all indecomposable projective objects in are of the form for some indecomposable or for some indecomposable and some ; and all indecomposable injective objects in are of the form for some indecomposable or for some indecomposable and some . Thus, for , is not Frobenius. For and any morphism of projectives, define by
[TABLE]
So for each projective object , we have that .
3. Hall algebras of -cyclic complexes and -term complexes
In this section, let , we study the relation between Hall algebras of -cyclic complexes and -term complexes.
Given objects , let be the subset consisting of those equivalence classes of short exact sequences with middle term .
Definition 3.1**.**
The Hall algebra of is the vector space over with basis elements , and with the multiplication defined by
[TABLE]
Remark 3.2**.**
Given objects , set
[TABLE]
It is clear that
[TABLE]
where
[TABLE]
By the Riedtmann–Peng formula [16, 14],
[TABLE]
Thus,
[TABLE]
In fact, in terms of alternative generators , the product takes the form
[TABLE]
which is the definition used, for example, in [18, 19].
Let (resp. ) be the Hall algebra of the abelian category (resp. ) as defined in Definition 3.1. Let (resp. ) be the subspace of (resp. ) spanned by the isomorphism classes of objects in (resp. ). Since (resp. ) is closed under extensions, (resp. ) is a subalgebra of the Hall algebra (resp. ).
Define to be the subspace of spanned by elements , where with . Denote by the complement space of in , which is spanned by elements satisfying that with . Namely, as vector spaces, we have the decomposition .
Lemma 3.3**.**
* is an ideal of .*
Proof..
We only prove for , since the case for can be similarly proved. Let and with , i.e., . For any , consider the short exact sequence
[TABLE]
where . Then we have that . Assume that , thus . We get that , since is injective. This is a contradiction, and thus we conclude that . That is, . Similarly, we prove that . Hence, is an ideal of . ∎
Given , we fix an object defined by
[TABLE]
where . In particular, for , .
Theorem 3.4**.**
* There exists a surjective homomorphism of algebras*
[TABLE]
* There exists an isomorphism of algebras*
[TABLE]
Proof..
Let be the linear map defined on basis elements by
[TABLE]
where . For any objects in , if , then if and only if ; assume that , in this case . Hence, is well defined.
Given objects , on the one hand, if or , i.e., or , thus . By Lemma 3.3 we have that , thus .
On the other hand, if and , then by (3.2), we obtain that
[TABLE]
where . Since , we obtain that
[TABLE]
It is clear that there exists a bijection between and , and . Hence,
[TABLE]
Therefore, is a homomorphism of algebras.
For any , take with and , then , and thus is surjective.
Since , induces a surjective homomorphism of algebras
[TABLE]
The injectivity of follows from the fact that sends the basis of to a basis of . ∎
4. Indecomposable objects in and
From now onwards, we always assume that is hereditary until the end of the entire paper. For each object , according to [4, Section 4.1], it has a minimal projective resolution111The notations and will be used throughout the paper.
[TABLE]
Moreover, we have the following well-known
Lemma 4.1**.**
Given , each projective resolution of is isomorphic to a resolution of the form
[TABLE]
for some and some minimal projective resolution
[TABLE]
We define objects for and for . By Lemma 4.1, we know that any two minimal projective resolutions of are isomorphic, so and are well defined up to isomorphism.
Now, let us give characterizations of indecomposable objects in and .
Lemma 4.2**.**
([8, Lemma 2.3])* For , the objects and , where , is indecomposable and is indecomposable, provide a complete set of indecomposable objects in . Moreover, all are the whole indecomposable projective-injective objects in .*
Lemma 4.3**.**
For , the objects , and , where , is indecomposable and is indecomposable, provide a complete set of indecomposable objects in . Moreover, all and are the whole indecomposable projective objects; all and are the whole indecomposable injective objects.
Proof..
We only prove the first statement, since the others have been given in the previous section. Take an arbitrary object in
[TABLE]
For each , we have the short exact sequence
[TABLE]
By the hereditary assumption, all the objects appearing in these sequences are projective. Thus the sequences split. That is, we can assume that for each . Writing as follows:
[TABLE]
where and elsewhere for , we obtain that is the direct sum of the following two objects
[TABLE]
and
[TABLE]
Similarly, is the direct sum of the following two objects
[TABLE]
[TABLE]
Repeating this process, we get that has a direct sum decomposition
[TABLE]
where ; for , with if , , and ; and with if , , and .
Noting for each the differential in is injective, we have the short exact sequence
[TABLE]
where . Then, by Lemma 4.1, for some . Set , since we have that . Therefore, we complete the proof.∎
5. Hall algebras of bounded complexes and derived Hall algebras
In order to study the Hall algebra of , by reformulating [22, Theorem 3.2] we first give a characterization of the Hall algebra of , and then relate it to the derived Hall algebra of . These are similar to the results given in [13, Section 5.2].
Let be the Hall algebra of the abelian category as defined in Definition 3.1. Let be the subspace of spanned by the isomorphism classes of objects in . Since is closed under extensions, is a subalgebra of the Hall algebra .
For objects , define
[TABLE]
Since is hereditary, it descends to give a bilinear form
[TABLE]
known as the Euler form.
Define the Hall algebra to be the same vector space as , but with the twisted multiplication defined by
[TABLE]
Given objects , there exists a positive integer such that . Since the global dimension of is , we obtain that there exists a positive integer such that for all , . That is, is locally homological finite. Thus, the Euler form of
[TABLE]
determined by
[TABLE]
is also well defined.
Lemma 5.1**.**
For any and , we have that
- (1)
* for any ;*
- (2)
;
- (3)
* for any ;*
- (4)
, if ;
- (5)
, if .
Proof..
For any , by [11, Lemma 3.1], we have that
[TABLE]
Thus, - can be easily proved. Noting that if , we prove ; Noting that if , we prove . ∎
Define the Hall algebra to be the same vector space as , but with the twisted multiplication
[TABLE]
For any and , since is projective-injective in , we easily obtain that
[TABLE]
for all .
Define the Hall algebra to be the localization of with respect to elements for all and . For each and , by writing for some , we define
[TABLE]
For any , it is easy to see that
[TABLE]
Moreover, for any we have that
[TABLE]
Given an object , for each , we define
[TABLE]
Let us reformulate [8, Proposition 4.4(2)] in the following
Proposition 5.2**.**
For each , there exists an embedding of algebras
[TABLE]
Remark 5.3**.**
Compared with the modified Ringel–Hall algebra studied in [13], the element corresponding to needs to be defined by appending the element to . Actually, in the modified Ringel–Hall algebra, the embedding above is immediate.
By Lemma 4.2, applying the arguments similar to those in the proof of [8, Proposition 4.4(3)], we obtain the following
Proposition 5.4**.**
* has a basis consisting of elements*
[TABLE]
where , , and for .
Given objects , we denote by the set
[TABLE]
and set
[TABLE]
Applying the arguments similar to those in the proof of [22, Theorem 3.2] together with Lemma 5.1, we obtain the following
Proposition 5.5**.**
The Hall algebra is generated by the elements in
[TABLE]
with the defining relations
[TABLE]
where , and .
Remark 5.6**.**
By Proposition 5.5 and [13, Proposition 5.3], we obtain that the Hall algebra is isomorphic to the modified Ringel–Hall algebra defined in [13]. Explicitly, there exists an isomorphism defined on generators by and .
For simplicity, we recall the twisted derived Hall algebra of in the form of generators and relations in the following
Proposition 5.7**.**
([20])* is an associative and unital -algebra generated by the elements in and the following relations*
[TABLE]
Now we reformulate [13, Theorem 5.5] in the following
Theorem 5.8**.**
There is an embedding of algebras
[TABLE]
defined on generators (with ) by
[TABLE]
Let be the subalgebra of generated by elements with and . By [13, Corollary 5.6], there is an isomorphism of algebras
[TABLE]
By [13, Corollary 5.7], we know that is invariant under derived equivalences. The inverse of is the homomorphism
[TABLE]
defined on generators (with ) by
[TABLE]
[TABLE]
where we have written elements as .
In fact, these results above are essentially the same as those given by Gorsky in [11, Theorem 4.2]. However, the explicit map between the Hall algebra of and the derived Hall algebra is not given there, besides, the twist therein used only in the Hall algebra of , not in the derived Hall algebra, not only involves , but also .
6. Hall algebras of -term complexes
Since the global dimension of is , the Euler form of
[TABLE]
determined by
[TABLE]
is well defined.
Define the Hall algebra to be the same vector space as , but with the twisted multiplication
[TABLE]
For any projective-injective object , we easily obtain that
[TABLE]
for all . Define the Hall algebra to be the localization of with respect to elements corresponding to projective-injective objects in . It is easy to see that is isomorphic to the group algebra of the Grothendieck group .
From now on, we let . As before, for each and , by writing for some , we define
[TABLE]
For any , it is easy to see that
[TABLE]
Moreover, for any we have that
[TABLE]
Given an object , for each , we define
[TABLE]
For each object , we define
[TABLE]
First of all, we give a basis in as follows:
Proposition 6.1**.**
* has a basis consisting of elements*
[TABLE]
where and for , and .
Proof..
It is similar to (5.3) that for any objects ,
[TABLE]
Thus, for and , noting that is injective in , we have that
[TABLE]
for some . By Lemma 4.3, we can easily complete the proof. ∎
Let us consider the shift functor . Clearly, is a fully faithful exact functor. Moreover, for any objects , we have that for all . That is, is an extremely faithful exact functor. By functorial properties of Hall algebras (cf. [19]), there exists an embedding of algebras . Thus, we have the following
Proposition 6.2**.**
*There exists an embedding of algebras \textstyle{\lambda:\mathcal{M}\mathcal{H}\,_{m}(\mathcal{A}\,)\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$$\textstyle{\mathcal{M}\mathcal{H}\,(\mathcal{A}\,).} *
Proof..
Let and be the natural homomorphisms of algebras. Since maps all elements in to the invertible elements in , by universal properties of localizations, we obtain that there is a unique homomorphism of algebras
[TABLE]
such that . Explicitly, , and for all , , and . Thus, the injectivity of follows from the fact that sends the basis of in Proposition 6.1 to a linearly independent set in . ∎
Combining Proposition 6.2 with Proposition 5.2, we obtain the following
Proposition 6.3**.**
For each , there exists an embedding of algebras
[TABLE]
Proof..
Taking gives the desired homomorphism. ∎
Combining Propositions 5.5, 6.1 and 6.2, we have the following
Proposition 6.4**.**
The Hall algebra is generated by the elements in
[TABLE]
with the defining relations
[TABLE]
where , and .
Let be the subalgebra of generated by elements with and .
Corollary 6.5**.**
There is an embedding of algebras
[TABLE]
Proof..
Taking gives the desired homomorphism. ∎
Remark 6.6**.**
is used in [9] to realize quantum cluster algebra with principal coefficients as its subquotient.
7. Integration maps associated to Hall algebras of
In this section, let be a finitary exact category of global dimension at most one. By [12], Definition 3.1 also applies to the exact category . We also have the Riedtmann–Peng formula in the Hall algebra of .
Let be the -algebra with a basis and the multiplication given by
[TABLE]
where is the Euler form on . According to [15], we recall the integration map on the Hall algebra as follows:
Proposition 7.1**.**
The integration map
[TABLE]
is a homomorphism of algebras.
Proof..
For reader’s convenience, we give the proof here. Given objects ,
[TABLE]
∎
Since the global dimension of is equal to one, we can apply the integration map in Proposition 7.1 to the Hall algebra of . Before doing this, we first give a characterization on the Grothendieck group of . By [1, Proposition 3.2], for each object M_{\bullet}=(\lx@xy@svg{\hbox{\raise 0.0pt\hbox{\kern 9.79582pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\kern 0.0pt\offinterlineskip\halign{\entry@#!@&&\entry@@#!@\cr&\crcr}}}\ignorespaces{\hbox{\kern-9.79582pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{\textstyle{M_{1}\ignorespaces\ignorespaces\ignorespaces\ignorespaces}}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 15.97412pt\raise 6.075pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.78612pt\hbox{\scriptstyle{d_{1}}}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 33.79582pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 33.79582pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{\textstyle{M_{2}}}}}}}}}\ignorespaces}}}}\ignorespaces)\in C^{2}(\mathscr{P}\,), we have the following injective resolution
[TABLE]
Let be all indecomposable projective objects in up to isomorphism. Fixing a minimal injective resolution of
[TABLE]
we define the dimension vector dim on objects in by setting
[TABLE]
By the dual of Lemma 4.1, we obtain that does not depend on the minimality of injective resolutions. Thus, by the dual of Horseshoe Lemma (cf. [5, Theorem 12.8]), we get the additivity of dim . That is, for any short exact sequence
[TABLE]
in , we have that .
Lemma 7.2**.**
The Grothendieck group is a free abelian group having as a basis the set
[TABLE]
and there exists a unique group isomorphism such that for each object in .
Proof..
For any object , taking an injective resolution of as (7.1), we obtain that in
[TABLE]
This shows that generates the group .
For any objects , it is clear that implies , since their minimal injective resolutions are isomorphic. Thus, the additivity of dim implies the existence of a unique group homomorphism such that for each object in . Since the image of the generating set under the homomorphism is the canonical basis of the free group , this set is -linearly independent in . It follows that is free and that is an isomorphism. ∎
Let be the bilinear form on obtained from the Euler form of by the isomorphism in Lemma 7.2. Define the quantum torus associated to the pair to be the -algebra with a basis and the multiplication given by
[TABLE]
Corollary 7.3**.**
The integration map
[TABLE]
is a homomorphism of algebras.
Remark 7.4**.**
(Further directions) Since the global dimension of is equal to one, we want to know whether the Hall algebra has a bialgebra structure, or say, whether Green’s formula on the Hall numbers of a hereditary abelian category holds in ;
Let be an acyclic quiver of vertices, and take to be the category of finite dimensional -modules. Whether we can use the integration map in Corollary 7.3 to relate the Hall algebra with quantum cluster algebras (with principal coefficients). Namely, Whether we can relate the integration map in Corollary 7.3 with the work in [9].
Acknowledgments
The author is grateful to Shiquan Ruan for his suggestion to relate Hall algebra of -cyclic complexes with Hall algebra of -term complexes. He also would like to thank Fan Xu for the conversations on integration maps.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] R. Bautista, The category of morphisms between projective modules, Comm. Algebra 32 (11) (2004), 4303–4331.
- 2[2] R. Bautista, M. Souto-Salorio and R. Zuazua, Almost split sequences for complexes of fixed size, J. Algebra 287 (2005), 140–168.
- 3[3] A. Berenstein and A. Zelevinsky, Quantum cluster algebras, Adv. Math. 195 (2) (2005), 405–455.
- 4[4] T. Bridgeland, Quantum groups via Hall algebras of complexes, Ann. Math. 177 (2013), 1–21.
- 5[5] T. Bühler, Exact Categories, Expo. Math. 28 (1) (2010), 1–69.
- 6[6] C. Chaio, I. Pratti and M. Souto-Salorio, On sectional paths in a category of complexes of fixed size, Algebr. Represent. Theor. 20 (2017), 289–311.
- 7[7] C. Chaio, A. Chaio and I. Pratti, On the type of a category of complexes of fixed size and the strong global dimension, Algebr. Represent. Theor., https://doi.org/10.1007/s 10468-018-9823-3.
- 8[8] Q. Chen and B. Deng, Cyclic complexes, Hall polynomials and simple Lie algebras, J. Algebra 440 (2015), 1–32.
