The number of independent Traces and Supertraces on the Symplectic Reflection Algebra $H_{1,\nu}(\Gamma \wr S_N)$
Extracted source text · arXiv 1711.09950
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% %
% S.E.Konstein, I.V.Tyutin %
% Symplectic Reflection Algebra %
% $H_{1,\nu}(\Gamma \wr $_N)$ %
% %
% %
% %
% %
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%% %%%
\newcommand{\ddatt}{ %%%
27-11-2017 %%%
} %% %%%
%%% %%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\documentclass[12pt]{article}
\usepackage{euscript,amsmath, amssymb, amsfonts}
\usepackage{color}
%\usepackage[cp1251]{inputenc}
%\usepackage[russian]{babel}
%\usepackage{hyperref}
%\usepackage{comments}
%\url{http://google.com}
\pagestyle{myheadings}
\hoffset=-25.0pt
\voffset=-2cm
\textwidth=16.5cm
\textheight=24cm
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% GENERAL DEFINITIONS
\newcommand{\ig}{\Big}
\newcommand{\igg}{\Bigg}
\newcommand{\sigmama}{ {{g}} }
\newcommand{\p}{\partial}
\newcommand{\R}{{\cal R}}
\newcommand{\HH}{ H_{1,\, \eta}(G) }
\newcommand{\GG}{{\Gamma \wr S_N}}
\newcommand{\GGG}{{\mathbb C[G]}}
\newcommand{\str}{sp }
\newcommand{\ce}{{\cal Z}^L }
\newcommand{\se}{{\cal Z}^S }
\newcommand{\LLL}{ L_{1,\, \eta}(G) }
\newcommand{\SSS}{ S_{1,\, \eta}(G) }
\newcommand{\be}{\begin{equation}}
\newcommand{\ee}{\end{equation}}
\newcommand{\bee}{\begin{eqnarray}}
\newcommand{\eee}{\end{eqnarray}}
\newcommand\nn{\nonumber \\}
\newcommand\n{ }
\newcommand\defeq{:=}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\newcounter{theorem}%
\newcommand{\theorem}{\par\refstepcounter{theorem}
{\bf Theorem
\arabic{section}.%
% \arabic{subsection}.
\arabic{theorem}. }}
\renewcommand\thetheorem{\thesection.%
%\arabic{subsection}.
\arabic{theorem}}
\makeatletter \@addtoreset{theorem}{section}
\newcounter{lemma}
\newcommand{\lemma}{\par\refstepcounter{theorem}
{\bf Lemma
\arabic{section}.%
% \arabic{subsection}.
\arabic{theorem}. }}
\renewcommand\thelemma{\thesection.
%\arabic{subsection}.
\arabic{theorem}}
\makeatletter \@addtoreset{lemma}{section}
\newcounter{proposition}
\newcommand{\proposition}{\par\refstepcounter{theorem}
{\bf Proposition
\arabic{section}.%
% \arabic{subsection}.
\arabic{theorem}. }}
\renewcommand\theproposition{\thesection.%
%\arabic{subsection}.
\arabic{theorem}}
\makeatletter \@addtoreset{proposition}{section}
\newcounter{corollary}
\newcommand{\corollary}{\par\refstepcounter{theorem}
{\bf Corollary
\arabic{section}.%
% \arabic{subsection}.
\arabic{theorem}. }}
\renewcommand\thecorollary{\thesection.\arabic{subsection}.\arabic{theorem}}
\makeatletter \@addtoreset{corollary}{section}
\newcounter{definition}
\newcommand{\definition}{\par\refstepcounter{theorem}
{\bf Definition
\arabic{section}.%
% \arabic{subsection}.
\arabic{theorem}. }}
\renewcommand\thedefinition{\thesection.%
%\arabic{subsection}.
\arabic{theorem}}
\makeatletter \@addtoreset{definition}{section}
\newenvironment{proof}[1][Proof]{\noindent\textsf{#1.\ }}
%{ \ \rule{0.5em}{0.5em}}
{\hfill {\small $\square$}}
\makeatletter \@addtoreset{equation}{section}
\def\theequation{\thesection.\arabic{equation}}
%\makeatother
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\begin{document}
\sloppy \title
{
The number of independent Traces and Supertraces on the Symplectic
Reflection Algebra $H_{1,\nu}(\Gamma \wr S_N)$ }
\author
{
S.E. Konstein%
\thanks{ I.E. Tamm Department of Theoretical Physics,
P.N. Lebedev Physical Institute, RAS
119991, Leninsky prosp., 53, Moscow, Russia}
\thanks{E-mail: konstein@lpi.ru}
%
,
I.V. Tyutin$^*$%
%\thanks{ I.E. Tamm Department of Theoretical Physics,
% P.N. Lebedev Physical Institute, RAS
% 119991, Leninsky prosp., 53, Moscow, Russia.}
\thanks{Tomsk State Pedagogical University, Kievskaya St. 60, 634061 Tomsk, Russia}
\thanks{E-mail: tyutin@lpi.ru} }
\date{
%\ddatt
}
\maketitle
\thispagestyle{empty}
\begin{abstract}
Symplectic reflection algebra $ H_{1, \,\nu}(G)$ has a
$T(G)$-dimensional space of traces whereas, when considered as a
superalgebra with a natural parity, it has an $S(G)$-dimensional
space of supertraces.
%
The values of $T(G)$ and $S(G)$ depend on the symplectic reflection
group $G$
%$G\leq Sp(2n,\mathbb C)$
and do not depend on the parameter $\nu$.
In this paper, the values $T(G)$ and $S(G)$ are explicitly calculated for the groups
$G= \Gamma \wr S_N$,
where $\Gamma$ is a finite subgroup of $ Sp(2,\mathbb C)$.
\end{abstract}
%\keywords{Symplectic Reflection Algebra; Cherednik algebra; trace, supertrace.}
%\ccode{2000 Mathematics Subject Classification: 17B80, 16W55}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Introduction}
\label{page}
Let $V:=\mathbb C^{2N}$, let $G\subset Sp(2N,\mathbb C)$ be a
finite group generated by symplectic reflections. In \cite{KT2}, it
was shown that Symplectic Reflection Algebra $H_{1,\nu}(G)$ has
$T(G)$ independent traces, where $T(G)$ is the number of conjugacy
classes of elements without eigenvalue $1$ belonging to the group
$G\subset Sp(2N)\subset End(V)$, and that the algebra
$H_{1,\nu}(G)$, considered as a superalgebra with a natural parity,
has $S(G)$ independent supertraces, where $S(G)$ is the number of
conjugacy classes of elements without eigenvalue $-1$ belonging to
$G\subset Sp(2N)\subset End(V)$. Hereafter, speaking about spectrum,
eigenvalues and eigenvectors, the rank of an element of the group
algebra $\mathbb C[G]$ of the group $G$, etc., we have in mind the
representation of the group algebra $\mathbb C[G]$ in the space $V$.
Besides, we denote all the units in groups, algebras, etc., by 1,
and $c\cdot 1$ by $c$ for any number $c$.
%Apart from a few groups,
There are two families of groups generated by symplectic
reflections, see \cite{GS} and also \cite{HW}, \cite{BG},
\cite{Gor}:
Family 1): $G$ is a complex reflection group acting on $\mathfrak H
\oplus \mathfrak H^*$, where $\mathfrak H$ is the space of
reflection representation. In this case, $G$ is a product of several
groups from the following set of Coxeter groups
\be\label{list}
A_n, \,\,B_n, \,\,C_n, \,\,D_n, \,\,E_n, \,\,F_n, \,\,G_2,
\,\,H_n, \,\,I_2(n).
\ee
Family 2): $G=\Gamma \wr S_N$, which means here
$G=\Gamma^N \rtimes S_N$
acting on $ (\mathbb
C^2)^N$, where $\Gamma$ is a finite subgroup of $Sp(2,\mathbb C)$.
For groups $G$ from the set (\ref{list}), the list of values $T(G)$ and $S(G)$ is given in
\cite{stek}.
\medskip
In this work, we give the values of $T(G)$ and $S(G)$ for the 2nd family.
Namely, we found the generating functions
\bee\label{th0}
t(\Gamma,x)&:= &\sum_{N=0}^{\infty}T(\Gamma \wr S_N)x^N,
\\
s(\Gamma,x)&:= &\sum_{N=0}^{\infty}S(\Gamma \wr S_N)x^N
\label{th00}
\eee
for each finite subgroup $\Gamma\subset
Sp(2,\mathbb C)$, see
Theorem \ref{th1}.
All needed definitions are given in the Section \ref{prel}; the
structure, conjugacy classes and
characteristic polynomials of the groups $\Gamma \wr S_N$ are described
in Section \ref{secG}.
To include the case $N=0$ in consideration in formulas (\ref{th0}) -- (\ref{th00}),
it is natural to set $\Gamma \wr S_0:= \{E\}$, where $\{E\}$ is the group
containing only one element $E$, and, since $\dim V=0$, set
$H_{1,\nu}(\Gamma \wr S_0):= \mathbb C[\{E\}]$.
Applying the definitions given in Section \ref{prel} to the algebra $H_{1,\nu}(\Gamma \wr S_0)$ we deduce that
a) if the algebra $H_{1,\nu}(\Gamma \wr S_0)= \mathbb C[\{E\}]$ is considered as superalgebra,
it has only a trivial parity $\pi\equiv 0$;
b) the algebra $H_{1,\nu}(\Gamma \wr S_0)= \mathbb C[\{E\}]$ has 1-dimensional space of traces
and 1-dimensional space of supertraces; these spaces coincide;
c) it is natural to set $T(\Gamma \wr S_0)= S(\Gamma \wr S_0)=1$;
d) the algebra $H_{1,\nu}(\Gamma \wr S_0)= \mathbb C[\{E\}]$
contains two
Klein operators (i.e., elements satisfying conditions
(\ref{k1}) -- (\ref{k3})), namely, $E$ and $-E$.
\section{Preliminaries}\label{prel}
\subsection{Traces}
Let ${\cal A}$ be an associative superalgebra with parity $\pi$.
All expressions of linear algebra are given for homogenous elements only
and are supposed to be extended to inhomogeneous elements via linearity.
A linear complex-valued function $str$ on ${\cal A}$ is called a {\it supertrace} if
$$str(fg)=(-1)^{\pi(f)\pi(g)}str(gf) \ \mbox{ for all } f,g\in {\cal A}.$$
A linear complex-valued function $tr$ on ${\cal A}$ is called a {\it trace} if
$$tr(fg)=tr(gf) \ \mbox{ for all } f,g\in {\cal A}.$$
The element $K\in \cal A$ is called a {\it Klein operator}, if
\begin{eqnarray}\label{k1}
&&\pi(K)=0,\\
\label{k2}
&&K^2=1,\\
&&Kf=(-1)^{\pi(f)} f K \qquad \mbox{ for all } f \in {\cal A}.
\label{k3}
\end{eqnarray}
Any Klein operator, if exists, establishes an isomorphism
between the space of traces on ${\cal
A}$ and the space of supertraces on ${\cal
A}$.
Namely, if $f\mapsto tr(f)$ is a trace, then $f\mapsto
tr(fK^{1+\pi(f)})$ is a supertrace, and if $f\mapsto str(f)$ is a
supertrace, then $f\mapsto str(fK^{1+\pi(f)})$ is a trace.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\subsection{Symplectic reflection group}
Let $V={\mathbb C}^{2N}$ be endowed with a non-degenerate
anti-symmetric
$Sp(2N)$-invariant bilinear form $\omega(\cdot,\cdot)$,
let the vectors $ e_i \in V$, where $i=1,\,...\,,\,2N$, constitute
a basis in $V$.
The matrix $(\omega_{ij}):=\omega(e_i,\,e_j)$ is anti-symmetric and non-degenerate.
Let $x^i$ be the coordinates of ${ x}\in V$, i.e.,
%
${x}= e_i\,x^i$. Then
%
$\omega({ x},\, y)=
\omega_{ij} x^i y^j$ for
any ${ x},\, y \in V$.
%
The indices $i$ are lowered and raised
by means of the forms $(\omega_{ij})$ and $(\omega^{ij})$, where $\omega_{ij}\omega^{kj}=\delta_i^k$.
\begin{definition}
The element $R\in Sp(2N)\subset End V$ is called a {\it symplectic reflection}, if
$rank(R-1)=2$.
\end{definition}
\begin{definition}
Any finite subgroup $G$ of $Sp(2N)$ generated by a set of symplectic reflections
is called a {\it symplectic reflection group}.
\end{definition}
In what follows, $G$ stands for a symplectic reflection group, and
$\R$ stands for the set of all symplectic reflections in $G$.
Let $R\in \R$.
Set
\bee
V_R &:=& Im (R-1) \,, \\
Z_R &:=& Ker (R-1) \,.
\eee
Clearly, $V_R$ and $Z_R$ are symplectically perpendicular, i.e., $\omega(V_R,\,Z_R)=0$,
and $V=V_R\oplus Z_R$.
%Hereafter the expression $U\oplus ^{\bot }W$
%denotes a direct sum whose summands are symplectically
%perpendicular to each other.
So, let
%
$x=x_{{\phantom \,}_{V_R}}+x_{{\phantom \,}_{Z_R}}$ for any $x\in V$,
where $x_{{\phantom \,}_{V_R}}\in V_R$ and $x_{{\phantom \,}_{Z_R}}\in Z_R$.
Set
\be
\omega_R(x,y):=\omega (x_{{\phantom \,}_{ V_R}},\,y_{{\phantom \,}_{V_R}}).
\ee
\subsection{Symplectic reflection algebra (following \cite{sra})}
Let $ {\mathbb C}[G] $ be the {\it group algebra} of $ {G} $, i.e., the
set of all linear combinations $\sum_{g\in {G} } \alpha_g \bar g$,
where $\alpha_g \in {\mathbb C}$.
If we were rigorists, we would
write $\bar g$
to distinguish $g$ considered as an element of $ {G} \subset End(V)$
from the same element $\bar g \in {\GGG}$
considered as an element of the group algebra.
The addition in $ {\GGG} $ is defined as follows:
$$
\sum_{g\in {G} } \alpha_g \bar g + \sum_{g\in {G} } \beta_g \bar g
= \sum_{g\in {G} } (\alpha_g + \beta_g) \bar g
$$
and the multiplication is defined by setting
$\overline {g_1\!}\,\, \overline {g_2\!} = \overline {g_1 g_2}$.
In what follows, however, we abuse notation and omit the bar sign over elements of the group algebra.
Let $\eta$ be a function on ${\R}$, i.e., a set of constants $\eta_R$ with $R\in\R$ such that
$\eta_{R_1}=\eta_{R_2}$ if $R_1$ and $R_2$ belong to
one conjugacy class of $ {G} $.
\begin{definition}
\label{defpage}
The algebra $H_{t,\eta}(G)$, where $t\in \mathbb C$, is an associative algebra with unit { 1};
it is the algebra
${\mathbb C}[V]$ of (noncommutative) polynomials in the elements of $V$ with coefficients
in the group algebra ${\mathbb C}[G]$ subject to the relations
\bee
g x&=&g(x) g %
\mbox{ for any } g\in {G}
\mbox{ and } x \in V,
\mbox{ where } g(x)= e_i g^i_j x^j
\mbox{ for }x=e_i x^i,\\
\label{rel}
\!\!\!\!\!\!\!\!\!\!\!\! [ x , y] &=& t \omega(x,y)
+
\sum_{R\in\R} \eta_R
\omega_R(x,y)R
\mbox{ for any $x,y\in V$}.
\eee
The algebra $H_{t,\eta}(G)$ is called a {\it symplectic reflection algebra},
see \cite{sra}.
\end{definition}
The commutation relations (\ref{rel}) suggest
to define the {\it parity} $\pi$ by setting:
\be\label{2.6}
\pi (x)=1,\ \pi (g)=0
\ \mbox{ for any }x\in V, \mbox{ and }g\in G,
\ee
enabling one to consider $H_{t,\eta}(G)$ as an associative {\it superalgebra}.
We consider the case $t\ne 0$ only, for any such $t$ it is equivalent to the case $t=1$.
Let ${\cal A}$ and ${\cal B}$ be superalgebras such that ${\cal A}$ is a ${\cal B}$-module.
We say that the superalgebra
${\cal A}\ast {\cal B}$ is a {\it crossed product} of ${\cal A}$ and ${\cal B}$ if
${\cal A}\ast {\cal B}={\cal A} \otimes {\cal B}$ as a superspace and
\[
(a_1\otimes b_1)\ast (a_2\otimes b_2)=a_1 b_1(a_2)\otimes b_1 b_2,
\]
see \cite{pass}.
The element $b_1(a_2)$ may include a sign factor imposed by the Sign Rule, see \cite{Del}, p. 45.
The (super)algebra $ H_{1, \,\eta}(G)$
is a deform of the crossed product
of the Weyl algebra $W_N$ and the group algebra
of a finite subgroup $G \subset Sp(2N)$
generated by symplectic reflections.
\subsection{
The number of independent traces and supertraces on the symplectic
reflection algebras }
%The following theorem was proved in \cite{KT2}:
\begin{theorem} (\cite{KT2})\label{main1}
{\it Let the symplectic reflection group $ {G} \subset End(V)$ have
$T_G$ conjugacy classes without eigenvalue $1$ and $S_G$ conjugacy
classes without eigenvalue $-1$.
Then the %super
algebra
$ H_{1, \,\eta}(G)$ has
$T(G)=T_G$ independent traces whereas $ H_{1, \,\eta}(G)$ considered
as a superalgebra, see (\ref{2.6}), has $S(G)=S_G$ independent
supertraces. }
\end{theorem}
\begin{proposition} Let $G$, $G_1$ and $G_2$ be symplectic reflection groups and $G=G_1 G_2$.
Then
$T(G_1 G_2)=T(G_1)T(G_2)$ and $S(G_1 G_2)=S(G_1)S(G_2)$.
\end{proposition}
\begin{proposition}\label{klein}
If there exists a $K\in G$ such that $K |_V=-1$,
then $K$ is a Klein operator.
\end{proposition}
\section{The group $\Gamma \wr S_N$}
\label{secG}
\subsection{Finite subgroups of $Sp(2,\mathbb C)$}
The complete list of the finite subgroups $\Gamma \subset Sp(2,\mathbb C)$
is as follows, see, e.g., \cite{raf}:
{%\footnotesize
\[
{{%
\begin{tabular}{|c|c|c|c|}
\hline {\bf $\Gamma$} & {\bf Order} & {\bf Presence of $-1$} &
\begin{tabular}{c}
{\bf The number of} \\
{\bf conjugacy classes}
{\bf $C(\Gamma)$}%
\end{tabular}
\\
\hline Cyclic group $\mathbf Z_n :=\mathbb{Z}/n\mathbb{Z}$ \ \ \ & $n$ & if \ $n$ is even & $n$ \\
\hline Binary dihedral group $\mathcal{D}_{n}\ \ $ & $4n$ & yes & $n+3$ \\
\hline Binary tetrahedral group $\mathcal{T}$ & $24$ & yes & $7$ \\
\hline Binary octahedral group $\mathcal{O}$ & $48$ & yes & $8$ \\
\hline Binary icosahedral group $\mathcal{I}$ & $120$ & yes & $9$ \\
\hline
\end{tabular}%
}}
\]%
}
It is easy to see that each of these groups, except
$\mathbf Z_{2k+1}$ with integer $k$, has $C(\Gamma) -1$
conjugacy classes without $+1$ in the spectrum and
has $C(\Gamma) -1$
conjugacy classes without $-1$ in the spectrum.
The group
$\mathbf Z_{2k+1}$ with integer $k$ has $C(\mathbf Z_{2k+1}) -1$
conjugacy classes without $+1$ in the spectrum and it
has $C(\mathbf Z_{2k+1})$
conjugacy classes without $-1$ in the spectrum.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\subsection{Symplectic reflections in $\Gamma\wr S_N$ (following \cite{EM})}
Let $V=\mathbb{C}^{2N}$ and let the symplectic form $\omega $ have the shape
\begin{equation}
\omega :=\left(
\begin{array}{cccc}
\varpi & & & \\
& \varpi & & \\
& & \ddots & \\
& & & \varpi
\end{array}%
\right),\ \text{ where }\varpi =\left(
\begin{array}{cc}
0 & 1 \\
-1 & 0%
\end{array}%
\right).
\end{equation}
The elements of the group $\Gamma \wr S_N$ have the form of $N
\times N$ block matrix with $2 \times 2$ blocks.
Consider the following elements of $\Gamma \wr S_N$
\bee
&& (D_{g,i})_{kl}:=\left\{
\begin{array}{l}
g, \ \ \text{ if } k=l=i, \\
1, \ \ \text{ if } k=l\ne i, \\
0, \ \ \text{ otherwise, }
\end{array}%
\right.
\label{D}
\\
&&(K_{ij})_{kl}:=\left\{
%\begin{array}
\begin{array}{ll}
\delta_{kl}, &\text{if } k,l\ne i, \ k,l\ne j, \\
\delta_{ki}\delta_{lj} + \delta_{kj}\delta_{li},&\text{otherwise, }\\
\end{array}%
\right.
\label{K}
\\
&& S_{g,ij}:=D_{g,i}D_{g^{-1},j}K_{ij},
\eee
%
where $i,j=1,...,N$, $i\ne j$, $1\ne g\in \Gamma$.
It is clear that $K_{ij}=K_{ji}$ and $ S_{g,ij}$ = $ S_{g^{-1},ji}$.
The complete set of symplectic reflections in $\Gamma\wr S_N$ consists of
$D_{g,i}$, $K_{ij}$
and $S_{g,ij}$, where
$1 \leqslant i < j \leqslant N$ and $1\ne g\in\Gamma$.
This set generates the group $\Gamma\wr S_N$.
The symplectic reflections $K_{ij}$ and $S_{g,ij}$ lie in one conjugacy class for all $i\ne j$ and $g\ne 1$;
the elements $D_{g,i}$ ($g\ne 1$) and $D_{h,j}$ ($h\ne 1$) lie in one conjugacy class if
$g$ and $h$ are conjugate in $\Gamma$.
So, the algebra $H_{1,\nu}(\Gamma^N \wr S_N)$ depends on $C(\Gamma)$ parameters $\nu$,
where $C(\Gamma)$ is the number of conjugacy classes in $\Gamma$ including the class $\{1\}$.
\subsection{Conjugacy classes}
Further,
the elements of the group $\Gamma \wr S_N$ can be represented in the form
$D\sigma$ where $D\in \Gamma^N$ is a diagonal
$N \times N$ block matrix, each block being a $2 \times 2$-matrix,
and $\sigma$ is $N \times N$ block matrix of permutation
each block being a $2 \times 2$-matrix.
The product has the form:
\[
(D_{1}\sigma _{1})(D_{2}\sigma _{2})=D_{3}\sigma _{3}
\]%
where $\sigma _{3}=\sigma _{1}\sigma _{2}$ and $D_{3}=D_{1}\sigma
_{1}D_{2}\sigma _{1}^{-1}.$
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Fix an element $g_{0}=D_{0}\sigma _{0}$. Since the permutation $%
\sigma _{0}$ is a product of cycles, there exists a~ permutation $\sigma
^{\prime }$ such that
\bee\label{dec}
\sigma ^{\prime }\sigma _{0}(\sigma ^{\prime })^{-1}=\left(
\begin{array}{cccc}
c_{1} & & & \\
& c_{2} & & \\
& & ... & \\
& & & c_{s}%
\end{array}%
\right),
%%
\text{ where $c_{k}$ are the cycles of length $L_{k}$, $\sum_{k}L_{k}=N$},
%\eee
%The cycles $c_{k}$ have the form%
%%%%
\\ \label{dec2}
c_{k}=\left(
\begin{array}{cccccc}
0 & 1 & 0 & 0 & ... & 0 \\
0 & 0 & 1 & 0 & ... & 0 \\
0 & 0 & 0 & 1 & ... & 0 \\
... & ... & ... & ... & ... & ... \\
0 & 0 & 0 & 0 & ... & 1 \\
1 & 0 & 0 & 0 & ... & 0%
\end{array}%
\right) .
\eee
The element $\sigma ^{\prime }D_{0}\sigma _{0}(\sigma ^{\prime })^{-1}$ has
the form%
\[
\sigma ^{\prime }D_{0}\sigma _{0}(\sigma ^{\prime })^{-1}=\left(
\begin{array}{cccc}
D_{1}c_{1} & & & \\
& D_{2}c_{2} & & \\
& & ... & \\
& & & D_{s}c_{s}%
\end{array}%
\right),
\]%
where $D_{k}$ is an $L_{k}\times L_{k}$ diagonal block matrix, each block being a $%
2\times 2$-matrix:%
\[
D_{k}=\left(
\begin{array}{cccc}
g_{1}^{k} & & & \\
& g_{2}^{k} & & \\
& & ... & \\
& & & g_{L_{k}}^{k}%
\end{array}%
\right) ,\text{ \ }g_{i}^{k}\in \Gamma .
\]%
Next, consider diagonal block matrices $%
H_{k}=diag(h_{1}^{k},h_{2}^{k},...,h_{L_{k}}^{k})$ and the elements
%$H_{k}D_{k}c_{k}H_{k}^{-1}$\ conjugated to $D_{k}c_{k}:$%
\[
H_{k}D_{k}c_{k}H_{k}^{-1}=\left(
\begin{array}{ccccc}
h_{1}^{k}g_{1}^{k}(h_{2}^{k})^{-1} & & & & \\
& h_{2}^{k}g_{2}^{k}(h_{3}^{k})^{-1} & & & \\
& & h_{3}^{k}g_{3}^{k}(h_{4}^{k})^{-1} & & \\
& & & ... & \\
& & & & h_{L_{k}}^{k}g_{L_{k}}^{k}(h_{1}^{k})^{-1}%
\end{array}%
\right) c_{k}.
\]
For any element $h_{1}^{k}\in \Gamma $, one can choose
\[
h_{2}^{k}=h_{1}^{k}g_{1}^{k}, \ h_{3}^{k}=h_{2}^{k}g_{2}^{k},\,
...,\ h_{L_{k}}^{k}=h_{L_{k}-1}^{k}g_{L_{k}-1}^{k}
\]
such that
\[
H_{k}D_{k}c_{k}H_{k}^{-1}=\left(
\begin{array}{ccccc}
1 & & & & \\
& 1 & & & \\
& & 1 & & \\
& & & ... & \\
& & & & h_{1}^{k}g_{1}^{k}g_{2}^{k}...g_{L_{k}}^{k}(h_{1}^{k})^{-1}%
\end{array}%
\right) c_{k}.
\]
So, each conjugacy class of $\Gamma \wr S_N$ is described by the set of cycles in the decomposition
(\ref{dec}), (\ref{dec2})
of $\sigma_0$, where each cycle is marked by some conjugacy class
of $\Gamma$.
The cycle of length $r$ marked by the conjugacy class $\alpha$ of $\Gamma$
with representative $g_\alpha\in \Gamma$ has the shape:
\be\label{marked-cyclec}
A_{\alpha,\,r}:=\left(
\begin{array}{ccccc}
1 & & & & \\
& 1 & & & \\
& & 1 & & \\
& & & ... & \\
& & & & g_\alpha%
\end{array}%
\right) c^{r}=\left(
\begin{array}{cccccc}
0 & 1 & 0 & 0 & ... & 0 \\
0 & 0 & 1 & 0 & ... & 0 \\
0 & 0 & 0 & 1 & ... & 0 \\
... & ... & ... & ... & ... & ... \\
0 & 0 & 0 & 0 & ... & 1 \\
g_\alpha & 0 & 0 & 0 & ... & 0%
\end{array}%
\right),
\ee
where $\alpha=1,...,C(\Gamma)$ and $r=1,2,...$;
%
$A_{\alpha,\,r}$ is an $r\times r$ block matrix, each block being a $2\times 2$ matrix.
So, each element $g\in\Gamma\wr S_N$ is conjugate to the element of the shape
\be\label{marked-cyc}
\left(
\begin{array}{ccccc}
A_{\alpha_1,\,r_1} & & & & \\
& A_{\alpha_2,\,r_2} & & & \\
& & ... & & \\
& & & & A_{\alpha_s,\,r_s}%
\end{array}%
\right)
\ee
It is convenient to describe the conjugacy class
of $\Gamma \wr S_N$ with representative (\ref{marked-cyc})
by the set of nonnegative integers $p_r^\alpha$,
where $r=1,\, 2,\,3,...\,$, and $\alpha=1,\,...\,,\,C(\Gamma)$, such that
\be\label{such}
\sum_{\alpha,\,r}r p_r^\alpha=N.
\ee
The value $p_r^\alpha$ for some conjugacy class is the number of cycles of length $r$
in the decomposition
(\ref{marked-cyc}) marked by the conjugacy class $\alpha$ of the group $\Gamma$.
The restriction (\ref{such}) can be omitted and can serve as definition of $N$ for each set of the numbers $p_r^\alpha$.
The number of conjugacy classes in $\Gamma \wr S_N$ is equal to
\be
C(\Gamma \wr S_N) = \sum_{p_r^\alpha:\, \sum_{\alpha,\,r}r p_r^\alpha=N} 1.
\ee
The generating function $c(\Gamma,x)$ of the number of conjugacy classes is defined as
\be
c(\Gamma,x):= \sum_{N=0}^\infty C(\Gamma \wr S_N)x^N
\ee
and is equal to
\be
c(\Gamma,x)=
\sum_{p_r^\alpha} x^{\sum_{\alpha,\,r}r p_r^\alpha}
= \sum_{p_r^\alpha =0}^\infty
\prod_{r=1}^\infty
\prod_{\alpha=1}^{C(\Gamma)}
(x^r)^{p_r^\alpha}=
\prod_{r=1}^\infty
\prod_{\alpha=1}^{C(\Gamma)}
\frac 1 {1-x^r}=\left(\Psi (x)\right)^{C(\Gamma)},
\ee
where $\Psi (x)$ is the Euler function
\be
\Psi (x):=\prod_{r=1}^\infty \frac 1 {1-x^r}
\ee
\subsection{Characteristic polynomials of conjugacy classes}
Before seeking the generating functions $t(\Gamma,x)$ and $s(\Gamma,x)$,
let us find the characteristic polynomial of the conjugacy class $g$ of $\Gamma\wr S_N$
identified by the set $p_r^\alpha$.
Let $P^{}_M(\lambda):=\det (M-\lambda)$ be the characteristic polynomial of the matrix $M$.
Then it is easy to see that
\[
P^{} _{A_{\alpha,\,r}}(\lambda )=\det (A_{\alpha,\,r}-\lambda )=\det
(g_\alpha-\lambda ^{r})=P^{} _{g_\alpha}(\lambda ^{r}),
\]
where the marked cycle $A_{\alpha,\,r}$ is defined by (\ref{marked-cyclec}).
Let $g\in\Gamma\wr S_N$ be defined by Eq. (\ref{marked-cyc}).
Now, it is easy to show that
\be\label{det}
P^{}_g(\lambda)=\det(g-\lambda)=
\prod_{i=1}^s \det( A_{\alpha_i,\, r_i}-\lambda ) =
\prod_{r,\alpha} (\det( g_\alpha-\lambda ^{r}) )^{p_r^\alpha},
\ee
%
if $g$ is a representative of the congugacy class in $\Gamma\wr S_N$ corresponding to the set
$p_r^\alpha$.
\begin{definition}
%\hskip -2mm
We call a conjugacy class {\it $t$-admissible}, if its representative $g\in \Gamma\wr S_N$
is such that $P^{}_g (1) \ne 0$.
\end{definition}
\begin{definition}
%\hskip -2mm
We call a conjugacy class {\it $s$-admissible}, if its representative $g\in \Gamma\wr S_N$
is such that $P^{}_g (-1) \ne 0$.
\end{definition}
%%%
\begin{definition}
%\hskip -2mm
We call a marked cycle $A_{\alpha,\, r}$, see Eq. (\ref{marked-cyclec}), {\it $t$-admissible},\\
if $P^{}_{A_{\alpha,\, r}} (1) \ne 0$.
\end{definition}
%%%
\begin{definition}
%\hskip -3mm
We call a marked cycle $A_{\alpha,\, r}$, see Eq. (\ref{marked-cyclec}), {\it $s$-admissible},\\
if $P^{}_{A_{\alpha,\, r}} (-1) \ne 0$.
\end{definition}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Recall that $g_\alpha\in\Gamma \subset Sp(2,\mathbb C)$, where $\Gamma$ is a finite group.
So $\det g_\alpha =1$ and the Jordan normal form of $g_\alpha$ is diagonal.
This implies that if $g_\alpha$ has $+1$ in its spectrum,
then $g_\alpha=1$ and if $g_\alpha$ has $-1$ in its spectrum,
then $g_\alpha=-1$.
These facts together with Eq. ({\ref{det}}) imply, in their turn,
the following two propositions:
%
\begin{proposition}\label{pr1}\it
The
conjugacy class %$g$
of $\Gamma\wr S_N$
identified by the set $p_r^\alpha$
is $t$-admissible
if and only if $g_\alpha\ne 1$ for all $\alpha$, $r$
with $p_r^\alpha\ne 0$.
\end{proposition}
\begin{proposition}\label{pr2}\it
The conjugacy class
of $\Gamma\wr S_N$
identified by the set $p_r^\alpha$
is $s$-admissible
if and only if
for any pair $\alpha,\,r$ such that $p_r^\alpha\ne 0$,
at least one of the next three conditions holds:
\begin{eqnarray*}
&&\text{ a) $g_\alpha\ne -1$ and $g_\alpha\ne 1$,}\\
&&\text{ b) $r$ is even and $g_\alpha = - 1$,}\\
&&\text{ c) $r$ is odd and $g_\alpha = 1$.}
\end{eqnarray*}
\end{proposition}
Note that the three sets of pairs $(r,\,\alpha)$ defined by the cases a), b), c)
in Proposition \ref{pr2} have empty pair-wise intersections.
Let $t_r(\Gamma)$ for $r=1,2...$ be equal to the number of different $\alpha$ such that $A_{\alpha,\, r}$
is $t$-admissible.
Evidently,
\be\label{tr}
t_r(\Gamma)=C(\Gamma)-1.
\ee
Analogously, let $s_r(\Gamma)$ for $r=1,2...$ be equal to the number of different $\alpha$ such that
$A_{\alpha,\, r}$
is $s$-admissible.
Evidently, if $\Gamma\ni -1$, then
\be\label{sr}
s_r(\Gamma)=C(\Gamma)-1.
\ee
and if $\Gamma\not\ni -1$, then
\be\label{srm}
s_{r}(\Gamma )=\left\{
\begin{array}{ll}
C(\Gamma )-1,&\text{ \ if }r\text{ is even}, \\
C(\Gamma ),&\text{ \ if }r\text{ is odd}.%
\end{array}%
\right.
\ee
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section{Combinatorial problem}
Consider the following simple combinatorial problem
(analogous problems are considered in \cite{hall}).
Suppose we have an unlimited supply of 1-gram colored weights for each of $n_1$ different colors,
an unlimited supply of 2-gram colored weights for each of $n_2$ different colors,
an unlimited supply of 3-gram colored weights for each of $n_3$ different colors, and so on.
Let $a^N_{n_1,\,...,\,n_k,\,...}$
be the number of opportunities to choose weights from our set of total mass $N$ grams.
The problem is to find generating function
\be
F_{n_1,...,n_k,...}(x) := \sum_{N=0}^\infty a^N_{n_1,\,...,\,n_k,\,...} x^N.
\ee
%
This problem is exactly the problem we discussed earlier. Namely, now we say \lq\lq $r$-gram weight"
instead of cycle of length $r$, and \lq\lq the number of different colors $n_r$" instead of
the number $t_r$ (\ref{tr}) or $s_r$ (\ref{sr})--(\ref{srm}) of different $\alpha$.
%the sets $t_r$ (\ref{tr}) or $s_r$ (\ref{sr})--(\ref{srm}).
\proposition\label{cp}
{\it
\be
F_{n_1+m_1,\, n_2+m_2,\,...,\,n_k+m_k,\,...}(x)
=
F_{n_1,\, n_2,\,...,\,n_k,\,...}(x) \cdot
F_{m_1,\, m_2,\,...,\,m_k,\,...}(x).
\ee
}
\begin{proof}
To prove this proposition, it suffices to note that
\be
a^N_{n_1+m_1,\, n_2+m_2,\,...,\,n_k+m_k,\,...}
=
\sum_{M=0}^N a^M_{n_1,\, n_2,\,...,\,n_k,\,...} \cdot
a^{N-M}_{m_1,\, m_2,\,...,\,m_k,\,...}.
\ee
\end{proof}
Introduce the functions
\be
f_i := F_{n^i_1,\, n^i_2,\,...,\,n^i_k,\,...}, \qquad \text{where $n^i_k=\delta^i_k$.}
\ee
Then
\be
f_i(x)=1+x^i+x^{2i}+x^{3i}+\,...\,= \frac 1 {1-x^i}.
\ee
The next theorem follows from Proposition \ref{cp}
\begin{theorem}\label{comb}
\be
F_{n_1,\, n_2,\,...,\,n_k,\,...}=\prod_{i=1}^\infty (f_i)^{n_i}.
\ee
\end{theorem}
The function $F_{1,\,1,\,1,\,...}=\Psi(x)$ is the well-known Euler function,
the generating function of the number of partitions of $N$ into the sum of positive integers.
\section{Generating functions $t(\Gamma)$ and $s(\Gamma)$}
\begin{theorem}\label{th1}\it Set
\bee
&&\Psi(x)=\prod_{i=1}^\infty \frac 1 {1-x^i} \qquad \text{ (Euler function)},\\
&&\Phi(x):=\prod_{k=0}^\infty \frac 1 {1-x^{2k+1}}.
\label{phi}
\eee
Let $T(\Gamma\wr S_N)$ be the dimension of the space of traces on $H_{1,\eta}(\Gamma\wr S_N)$
and let
$S(\Gamma\wr S_N)$ be the dimension of the space of supertraces on $H_{1,\eta}(\Gamma\wr S_N)$
considered as a superalgebra.
Let
\[
t(\Gamma,x) :=\sum_{N=0}^\infty T(\Gamma\wr S_N)x^N\text{~~
and~~}s(\Gamma,x) :=\sum_{N=0}^\infty S(\Gamma\wr S_N)x^N.
\]
\noindent Then
\bee
&&\text{ $t(\Gamma,x)=(\Psi(x))^{C(\Gamma)-1}$},\\
&&\text{ $s(\Gamma,x)=(\Psi(x))^{C(\Gamma)-1}$,\ \ \ \ \ \ \ \
if $\Gamma \ne \mathbf Z_{2k+1}$},\\
&&\text{ $s(\Gamma,x)=(\Psi(x))^{C(\Gamma)-1}\Phi(x) $,\ \
if $\Gamma = \mathbf Z_{2k+1}$.}
\eee
\end{theorem}
\begin{proof}
To prove Theorem \ref{th1}, we apply Theorem \ref{comb}
to the numbers (\ref{tr}) -- (\ref{srm}) of admissible conjugacy classes.
It is clear that
\be
t(\Gamma)=F_{t_1(\Gamma),\,t_2(\Gamma),\,t_3(\Gamma),\,...}=\Psi^{C(\Gamma)-1},
\ee
\be
s(\Gamma)=F_{s_1(\Gamma),\,s_2(\Gamma),\,s_3(\Gamma),\,...}
=
\left\{
\begin{array}{ll}
\Psi^{C(\Gamma )-1},&\text{ \ if }\Gamma \ni -1, \\
\Psi^{C(\Gamma )-1}\Phi,&\text{ \ if }\Gamma \not\ni -1.
\end{array}
\right.
\ee
\end{proof}
Observe that $\Phi(x)=\sum_{i=0}^\infty O_N x_N$, where $O_N$ is the number of
partitions of $N$ into the sum of odd positive integers, and $O_N$ coinsides with
the number of independent supertraces on $H_{1\nu}(S_N)$, see \cite{KV}.
\subsection{Inequality theorem}
\begin{theorem}\label{in2}
{\it Let $G=\Gamma \wr S_N$.
For each positive integer $N$, the following statements hold:
\bee
&&S(G)>0,\\
&&S(G)\geqslant T(G), \\
&& S(G) = T(G) \text{ if and only if } H_{1,\nu}(G) \text{ contains a Klein operator.}
\eee
}
\end{theorem}
Literally the same statements were proved for the Coxeter groups $G$ in \cite{stek},
and hence these statements hold for the product of any finite number of groups from
Family 1) and Family 2) defined on page \ref{page}.
\begin{proof}
Let $\Gamma \ne \mathbf Z_{2k+1}$.
Since each finite group $\Gamma\in Sp(2N,\mathbb C)$, except $\Gamma= \mathbf Z_{2k+1}$, contains $-1$,
the group $\Gamma\wr S_N$ contains Klein operator $K = \prod_{i=1}^N D_{-1,\,i}$.
There is no Klein operator in $H_{1,\nu}(\mathbf Z_{2k+1}\wr S_N)$ since for this algebra,
$S(\mathbf Z_{2k+1}\wr S_N)>T(\mathbf Z_{2k+1}\wr S_N)$, as it follows from
Theorem \ref{th1}.\end{proof}
%Analogous theorem for Coxeter groups was proved in \cite{stek}.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\section*{Acknowledgments}
The authors (S.K. and I.T.) are grateful to Russian Fund for Basic Research
(grant No.~${\text{17-02-00317}}$)
for partial support of this work.
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
\begin{thebibliography}{00}
\bibitem{Del} Bernstein J., \lq\lq Notes on supersymmetry\rq\rq.
In: Deligne P., Etingof P., Freed D., Jeffrey L., Kazhdan D., Morgan~J.,
Morrison~D., Witten~E. (eds.) \emph{Quantum fields and strings:
a~course for mathematicians}. Vol.~1. Material from the Special Year
on Quantum Field Theory held at the Institute for Advanced Study,
Princeton, NJ, 1996--1997. American Mathematical Society,
Providence, RI; Institute for Advanced Study (IAS), Princeton, NJ,
1999. Vol.~1: xxii+723~pp.
\bibitem{BG} Brown K.A., Gordon I.:
\lq\lq Poisson orders, symplectic reflection algebras and representation theory\rq\rq,
J. Reine Angew. Math. 559, 193216 (2003)
\bibitem{sra} P. Etingof and V. Ginzburg,
\lq\lq Symplectic reflection algebras, Calogero--Moser space, and
deformed Harish--Chandra homomorphism\rq\rq,
Inv. Math. 147 (2002), 243 -- 348.
\bibitem{EM} P. Etingof and S. Montarani,
\lq\lq Finite-dimensional representations of symplectic reflection algebras
associated to wreath products\rq\rq,
Representation Theory,
An Electronic Journal of the American Mathematical Society Volume 9,
Pages 457467 (July 21, 2005) S 1088-4165(05)00288-8.
\bibitem{Gor} Gordon, I.: \lq\lq Symplectic reflection algebras\rq\rq.
BMC talk (2005),\\
http://www.maths.ed.ac.uk/~igordon/BMC.pdf
Gordon, I.: Symplectic reflection algebras.
In: Trends in Representation Theory of Algebras and Related Topics, EMS Ser. Congr. Rep.,
Eur. Math. Soc., Zurich, pp. 285347 (2008)
\bibitem{GS} R. Guralnick and J. Saxl,
\lq\lq Generation of finite almost simple groups by conjugates\rq\rq, J. Algebra, 268
(2003), 519571.
\bibitem{hall} M.~Hall, {\it Combinatorial theory}, 2nd edition,
Wisley, New York (1986), 464 pp.
%\bibitem{lan}
S.K.~Lando, {\it Lectures on generating functions}.
Series: Student mathematical library, v. 23,
American Mathematical Society in Providence, RI, (2003), 144 pp.
\bibitem{HW} C.W. Huffman and D. Wales, \lq\lq Linear groups containing an element with an eigenspace of codimension
two\rq\rq, in: Proceedings of the Conference on Finite Groups, Univ. Utah, Park City, Utah, 1975,
Academic Press, New York, 1976, 425429.
\bibitem{stek} S.E. Konstein and R. Stekolshchik,
\lq\lq Klein operator and the Number of Traces and Supertraces
on the Superalgebra of Observables of Rational Calogero Model
based on the Root System\rq\rq,
Journal of Nonlinear Mathematical Physics,
Vol. 20:2 (2013), 295 -- 308.
% arXiv:0811.2487;
% arXiv:1212.0508.
\bibitem{KT2} S.E.~Konstein, I.V.~Tyutin,
\lq\lq The number of independent traces and supertraces on symplectic
reflection algebras\rq\rq, Journal of Nonlinear Mathematical Physics,
21:3, (2014) 308 -- 335; arXiv:1308.3190
\bibitem{KV} S.E.~Konstein and M.A.~Vasiliev,
``Supertraces on the Algebras of Observables
of the Rational Calogero Model with Harmonic Potential'',
J.~Math.~Phys. {\bf 37} (1996) 2872.
\bibitem{pass} D.S. Passman, Infinite Crossed Products,
Pure and Applied Math vol. 135, Academic Press, San Diego, 1989.
\bibitem{raf} R.~Stekolshchik,
{\it Notes on Coxeter Transformatios and the McKay Correspondence}.
Series: Springer Monographs in Mathematics 2008, XX, 239 pp.
\end{thebibliography}
\end{document}