Real-time collaboration for Jupyter Notebooks, Linux Terminals, LaTeX, VS Code, R IDE, and more,
all in one place.
Real-time collaboration for Jupyter Notebooks, Linux Terminals, LaTeX, VS Code, R IDE, and more,
all in one place.
| Download
GAP 4.8.9 installation with standard packages -- copy to your CoCalc project to get it
Project: cocalc-sagemath-dev-slelievre
Views: 418346\Chapter{The catalog of almost crystallographic groups}12This chapter introduces the access functions to the catalog of33- and 4-dimensional crystallographic groups. This catalog is an4electronic version of the classification obtained in \cite{KD}.56%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%7\Section{Rational matrix groups}89The following three main functions are available to access the library10of almost crystallographic groups as rational matrix groups.1112\> AlmostCrystallographicGroup( <dim>, <type>, <parameters> )13\> AlmostCrystallographicDim3( <type>, <parameters> )14\> AlmostCrystallographicDim4( <type>, <parameters> )1516<dim> is the dimension of the required group. Thus <dim> must be17either 3 or 4. The inputs <type> and <parameters> are used to define18the desired group as described in \cite{KD}. We outline the possible19choices for <type> and <parameters> here briefly. A more extended20description is given later in Section "More about almost crystallographic21groups" or can be obtained from \cite{KD}.2223<type> specifies the type of the required group. There are 17 types24in dimension 3 and 95 types in dimension 4. The input <type> can either25be an integer defining the position of the desired type among all types;26that is, in this case <type> is a number in [1..17] in dimension 3 or a27number in [1..95] in dimension 4. Alternatively, <type> can be a string28defining the desired type. In dimension 3 the possible strings are29`"01"', `"02"', $\ldots$, `"17"'. In dimension 4 the possible strings30are listed in the list `ACDim4Types' and thus can be accessed from \GAP.3132<parameters> is a list of integers. Its length depends on the type of33the chosen group. The lists `ACDim3Param' and `ACDim4Param' contain34at position $i$ the length of the parameter list for the type number $i$.35Every list of integers of this length is a valid <parameter> input.36Alternatively, one can input `false' instead of a parameter list. Then37\GAP\ will chose a random parameter list of suitable length.3839\beginexample40gap> G := AlmostCrystallographicGroup( 4, 50, [ 1, -4, 1, 2 ] );41<matrix group of size infinity with 5 generators>42gap> DimensionOfMatrixGroup( G );43544gap> FieldOfMatrixGroup( G );45Rationals46gap> GeneratorsOfGroup( G );47[ [ [ 1, 0, -1/2, 0, 0 ], [ 0, 1, 0, 0, 1 ], [ 0, 0, 1, 0, 0 ],48[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ],49[ [ 1, 1/2, 0, 0, 0 ], [ 0, 1, 0, 0, 0 ], [ 0, 0, 1, 0, 1 ],50[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ],51[ [ 1, 0, 0, 0, 0 ], [ 0, 1, 0, 0, 0 ], [ 0, 0, 1, 0, 0 ],52[ 0, 0, 0, 1, 1 ], [ 0, 0, 0, 0, 1 ] ],53[ [ 1, 0, 0, 0, 1 ], [ 0, 1, 0, 0, 0 ], [ 0, 0, 1, 0, 0 ],54[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ],55[ [ 1, -4, 1, 0, 1/2 ], [ 0, 0, -1, 0, 0 ], [ 0, 1, 0, 0, 0 ],56[ 0, 0, 0, 1, 1/4 ], [ 0, 0, 0, 0, 1 ] ] ]57gap> G.1;58[ [ 1, 0, -1/2, 0, 0 ], [ 0, 1, 0, 0, 1 ], [ 0, 0, 1, 0, 0 ],59[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ]60gap> ACDim4Types[50];61"076"62gap> ACDim4Param[50];63464\endexample6566%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%67\Section{Polycyclically presented groups}6869All the almost crystallographic groups considered in this package are70polycyclic. Hence they have a polycyclic presentation and this can be71used to facilitate efficient computations with the groups. To obtain the72polycyclic presentation of an almost crystallographic group we supply the73following functions. Note that the share package {\sf Polycyclic} must be74installed to use these functions.7576\> AlmostCrystallographicPcpGroup( <dim>, <type>, <parameters> )77\> AlmostCrystallographicPcpDim3( <type>, <parameters> )78\> AlmostCrystallographicPcpDim4( <type>, <parameters> )7980The input is the same as for the corresponding matrix group functions.81The output is a pcp group isomorphic to the corresponding matrix group.82An explicit isomorphism from an almost crystallographic matrix group83to the corresponding pcp group can be obtained by the following function.8485\> IsomorphismPcpGroup( <G> )8687We can use the polycyclic presentations of almost crystallographic88groups to exhibit structure information on these groups. For example,89we can determine their Fitting subgroup and ask group-theoretic90questions about this nilpotent group. The factor $G / Fit(G)$ of an91almost crystallographic group $G$ is called *holonomy group*. We92provide access to this factor of a pcp group via the following93functions. Let $G$ be an almost crystallographic pcp group.9495\> HolonomyGroup( <G> )96\> NaturalHomomorphismOnHolonomyGroup( <G> )9798The following example shows applications of these functions.99100\beginexample101gap> G := AlmostCrystallographicPcpGroup( 4, 50, [ 1, -4, 1, 2 ] );102Pcp-group with orders [ 4, 0, 0, 0, 0 ]103gap> Cgs(G);104[ g1, g2, g3, g4, g5 ]105106gap> F := FittingSubgroup( G );107Pcp-group with orders [ 0, 0, 0, 0 ]108gap> Centre(F);109Pcp-group with orders [ 0, 0 ]110gap> LowerCentralSeries(F);111[ Pcp-group with orders [ 0, 0, 0, 0 ], Pcp-group with orders [ 0 ],112Pcp-group with orders [ ] ]113gap> UpperCentralSeries(F);114[ Pcp-group with orders [ 0, 0, 0, 0 ], Pcp-group with orders [ 0, 0 ],115Pcp-group with orders [ ] ]116gap> MinimalGeneratingSet(F);117[ g2, g3, g4 ]118119gap> H := HolonomyGroup( G );120Pcp-group with orders [ 4 ]121gap> hom := NaturalHomomorphismOnHolonomyGroup( G );122[ g1, g2, g3, g4, g5 ] -> [ g1, identity, identity, identity, identity ]123gap> U := Subgroup( H, [Pcp(H)[1]^2] );124Pcp-group with orders [ 2 ]125gap> PreImage( hom, U );126Pcp-group with orders [ 2, 0, 0, 0, 0 ]127\endexample128129%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%130\Section{More about the type and the defining parameters}131132Each group from this library knows that it is almost crystallographic and,133additionally, it knows its type and defining parameters.134135\> AlmostCrystallographicInfo( <G> )136137This attribute is set for groups from the library only. It is not possible138at current to determine the type and the defining parameters for an arbitrary139almost crystallographic groups which is not defined by the library access140functions.141142\beginexample143gap> G := AlmostCrystallographicGroup( 4, 70, false );144<matrix group of size infinity with 5 generators>145gap> IsAlmostCrystallographic(G);146true147gap> AlmostCrystallographicInfo(G);148rec( dim := 4, type := 70, param := [ 1, -4, 1, 2, -3 ] )149\endexample150151\beginexample152gap> G := AlmostCrystallographicPcpGroup( 4, 70, false );153Pcp-group with orders [ 6, 0, 0, 0, 0 ]154gap> IsAlmostCrystallographic(G);155true156gap> AlmostCrystallographicInfo(G);157rec( dim := 4, type := 70, param := [ -3, 2, 5, 1, 0 ] )158\endexample159160We consider the types of almost crystallographic groups in more detail. The161almost crystallographic groups in dimensions 3 and 4 fall into three families162163\beginlist164\item{(1)} 3-dimensional almost crystallographic groups.165\item{(2)} 4-dimensional almost crystallographic groups with a166Fitting subgroup of class 2.167\item{(3)} 4-dimensional almost crystallographic groups with a168Fitting subgroup of class 3.169\endlist170171These families are split up further into subfamilies in \cite{KD} and to172each subfamily is assigned a type; that is, a string which is used to173identify the subfamily. As mentioned above, for the 3-dimensional almost174crystallographic groups the type is a string representing the numbers from1751 to 17, i.e. the available types are `"01"', `"02"', $\ldots$, `"17"'.176177For the 4-dimensional almost crystallographic groups with a Fitting subgroup178of class 2 the type is a string of 3 or 4 characters. In general, a string of1793 characters representing180the number of the table entry in \cite{KD} is used. So possible types are181`"001"', `"002"', $\ldots$. The reader is warned however that not all182possible numbers are used, e.g.\ there are no groups of type `"016"'. Also,183the types do not appear in their natural order in \cite{KD}. Moreover, for184certain numbers there is more than one family of groups listed in \cite{KD}.185For example, the 3 families of groups corresponding to number 19 on pages186179-180 of \cite{KD} have types `"019"', `"019b"' and `"019c"' (the order is187the one given in \cite{KD}).188189For the last category of groups, the 4-dimensional almost crystallographic190groups with a Fitting subgroup of class 3, the type is a string of 2 or 3191characters, where the first character is always the letter `"B"'. This `"B"'192is followed by the number of the table entry as found in \cite{KD},193eventually followed by a `"b"' or `"c"' as in the previous case.194195For each type of almost crystallographic group contained in the library196there exists a function taking a parameter list as input and returning197the desired matrix or pcp group. These functions can be accessed198from \GAP\ using the lists `ACDim3Funcs', `ACDim4Funcs', `ACPcpDim3Funcs'199and `ACPcpDim4Funcs' which consist of the corresponding functions.200201Although we include these direct access functions here for completeness,202we note that the user should in general use the higher-level functions203introduced above to obtain almost crystallographic groups from the204library. In particular, these low-level access functions return matrix205or pcp groups, but the almost crystallographic info flags will not be206attached to them.207208\beginexample209gap> ACDim3Funcs[15];210function( k1, k2, k3, k4 ) ... end211gap> ACDim3Funcs[15](1,1,1,1);212<matrix group with 5 generators>213gap> ACPcpDim3Funcs[1](1);214Pcp-group with orders [ 0, 0, 0 ]215\endexample216217%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%218\Section{The electronic versus the printed library}219220The package `aclib' can be considered as the electronic version of221Chapter 7 of \cite{KD}. In this section we outline the relationship222between the library presented in this manual and the printed version223in \cite{KD}. First we consider an example. At page 175 of \cite{KD},224we find the following groups in the table starting with entry ``13''.225226\medskip22713. $Q=P2/c$228$$229\matrix{ E:\;\langle a,b,c,d,\alpha,\beta\;\|\; & {\,[b,a]=1}\hskip 1.61cm230{[d,a]=1}231\hfill & \rangle \cr232& \matrix{ {[c,a]=d^{2 k_1}}\hfill & {[d,b]=1}\hfill \cr233{[c,b]=1}\hfill & {[d,c]=1}\hfill \cr234\alpha a=a^{-1}\alpha d^{k_{2}}\hfill & \alpha^2=d^{k_3}\hfill \cr235\alpha b=b\alpha \hfill & \alpha d= d \alpha\hfill \cr236\alpha c=c^{-1}\alpha d^{-2 k_6}\hfill & \cr237\beta a=a^{-1}\beta d^{k_1+k_2} \hfill & \beta^2=d^{k_5}\hfill \cr238\beta b=b^{-1}\beta d^{k_4}\hfill & \beta d= d \beta\hfill \cr239\beta c=c^{-1}\beta d^{-2 k_6}\hfill & \alpha \beta=c\beta\alpha d^{k_6}240\hfill } & }241$$242243$$\lambda(\alpha)=\left(\matrix{2441& \frac{k_1}{2}+k_2 & 0 & -2 k_6 & \frac{k_3}{2}+\frac{k_6}{2} \cr2450& -1 & 0 & 0 & 0 \cr2460 & 0& 1 & 0 & 0 \cr2470 & 0 & 0 & -1 & \frac12\cr2480 & 0 & 0 & 0 & 1249}\right)250\;\;\lambda(\beta)=\left(\matrix{2511& k_1+k_2 & k_4 & -2 k_6 & \frac{k_5}{2} \cr2520& -1 & 0 & 0 & 0 \cr2530 & 0& -1 & 0 & 0 \cr2540 & 0 & 0 & -1 & 0\cr2550 & 0 & 0 & 0 & 1256}\right)257$$258259$$H^2(Q,\Z{})=\Z{}\oplus(\Z{}_2)^4=\Z^{6}/A,$$260$$A=\{(k_1,\ldots,k_6)\|261k_1=0,\;k_2,\ldots, k_5\in 2\Z{},\;k_6\in\Z\}$$262AB-groups:263264$\forall k>0,\;k\equiv 0\bmod 2,\;<(k,0,1,0,1,0)>$265266\medskip267268The number ``13'' at the beginning of this entry is the type of the269almost crystallographic group in this library. This family of groups270with type 13 depends on 6 parameters $k_1, k_2, \ldots, k_6$ and these271are the <parameters> list in this library. The rational matrix272representation in \GAP\ corresponds exactly to the printed version in273\cite{KD} where it is named $\lambda$. In the example below, we consider274the group with parameters $(k_1,k_2,k_3,k_4,k_5,k_6)=(8,0,1,0,1,0)$.275276\beginexample277gap> G:=AlmostCrystallographicDim4("013",[8,0,1,0,1,0]);278<matrix group with 6 generators>279gap> G.5;280[ [ 1, 4, 0, 0, 1/2 ], [ 0, -1, 0, 0, 0 ], [ 0, 0, 1, 0, 0 ],281[ 0, 0, 0, -1, 1/2 ], [ 0, 0, 0, 0, 1 ] ]282gap> G.6;283[ [ 1, 8, 0, 0, 1/2 ], [ 0, -1, 0, 0, 0 ], [ 0, 0, -1, 0, 0 ],284[ 0, 0, 0, -1, 0 ], [ 0, 0, 0, 0, 1 ] ]285\endexample286287For a 4-dimensional almost crystallographic group the matrix group is288built up such that $\{ a, b, c, d, \alpha, \beta, \gamma \}$ as described289in \cite{KD} forms the defining generating set of $G$. For certain types290the elements $\alpha$, $\beta$ or $\gamma$ may not be present.291Similarly, for a 3-dimensional group we have the generating set $\{ a, b,292c, \alpha, \beta \}$ and $\alpha$ and $\beta$ may be absent.293294\bigskip295296To obtain a polycyclic generating sequence from the defining generators297of the matrix group we have to order the elements in the generating set298suitably. For this purpose we take the subsequence of $(\gamma, \beta,299\alpha, a, b, c, d)$ of those generators which are present in the300defining generating set of the matrix group. This new ordering of the301generators is then used to define a polycyclic presentation of the given302almost crystallographic group.303304305306