CoCalc Logo Icon
StoreFeaturesDocsShareSupportNewsAboutSign UpSign In

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

Path: gap4r8 / pkg / aclib / doc / aclib.tex
Views: 418346
1
\Chapter{The catalog of almost crystallographic groups}
2
3
This chapter introduces the access functions to the catalog of
4
3- and 4-dimensional crystallographic groups. This catalog is an
5
electronic version of the classification obtained in \cite{KD}.
6
7
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
8
\Section{Rational matrix groups}
9
10
The following three main functions are available to access the library
11
of almost crystallographic groups as rational matrix groups.
12
13
\> AlmostCrystallographicGroup( <dim>, <type>, <parameters> )
14
\> AlmostCrystallographicDim3( <type>, <parameters> )
15
\> AlmostCrystallographicDim4( <type>, <parameters> )
16
17
<dim> is the dimension of the required group. Thus <dim> must be
18
either 3 or 4. The inputs <type> and <parameters> are used to define
19
the desired group as described in \cite{KD}. We outline the possible
20
choices for <type> and <parameters> here briefly. A more extended
21
description is given later in Section "More about almost crystallographic
22
groups" or can be obtained from \cite{KD}.
23
24
<type> specifies the type of the required group. There are 17 types
25
in dimension 3 and 95 types in dimension 4. The input <type> can either
26
be an integer defining the position of the desired type among all types;
27
that is, in this case <type> is a number in [1..17] in dimension 3 or a
28
number in [1..95] in dimension 4. Alternatively, <type> can be a string
29
defining the desired type. In dimension 3 the possible strings are
30
`"01"', `"02"', $\ldots$, `"17"'. In dimension 4 the possible strings
31
are listed in the list `ACDim4Types' and thus can be accessed from \GAP.
32
33
<parameters> is a list of integers. Its length depends on the type of
34
the chosen group. The lists `ACDim3Param' and `ACDim4Param' contain
35
at position $i$ the length of the parameter list for the type number $i$.
36
Every list of integers of this length is a valid <parameter> input.
37
Alternatively, one can input `false' instead of a parameter list. Then
38
\GAP\ will chose a random parameter list of suitable length.
39
40
\beginexample
41
gap> G := AlmostCrystallographicGroup( 4, 50, [ 1, -4, 1, 2 ] );
42
<matrix group of size infinity with 5 generators>
43
gap> DimensionOfMatrixGroup( G );
44
5
45
gap> FieldOfMatrixGroup( G );
46
Rationals
47
gap> GeneratorsOfGroup( G );
48
[ [ [ 1, 0, -1/2, 0, 0 ], [ 0, 1, 0, 0, 1 ], [ 0, 0, 1, 0, 0 ],
49
[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ],
50
[ [ 1, 1/2, 0, 0, 0 ], [ 0, 1, 0, 0, 0 ], [ 0, 0, 1, 0, 1 ],
51
[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ],
52
[ [ 1, 0, 0, 0, 0 ], [ 0, 1, 0, 0, 0 ], [ 0, 0, 1, 0, 0 ],
53
[ 0, 0, 0, 1, 1 ], [ 0, 0, 0, 0, 1 ] ],
54
[ [ 1, 0, 0, 0, 1 ], [ 0, 1, 0, 0, 0 ], [ 0, 0, 1, 0, 0 ],
55
[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ],
56
[ [ 1, -4, 1, 0, 1/2 ], [ 0, 0, -1, 0, 0 ], [ 0, 1, 0, 0, 0 ],
57
[ 0, 0, 0, 1, 1/4 ], [ 0, 0, 0, 0, 1 ] ] ]
58
gap> G.1;
59
[ [ 1, 0, -1/2, 0, 0 ], [ 0, 1, 0, 0, 1 ], [ 0, 0, 1, 0, 0 ],
60
[ 0, 0, 0, 1, 0 ], [ 0, 0, 0, 0, 1 ] ]
61
gap> ACDim4Types[50];
62
"076"
63
gap> ACDim4Param[50];
64
4
65
\endexample
66
67
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
68
\Section{Polycyclically presented groups}
69
70
All the almost crystallographic groups considered in this package are
71
polycyclic. Hence they have a polycyclic presentation and this can be
72
used to facilitate efficient computations with the groups. To obtain the
73
polycyclic presentation of an almost crystallographic group we supply the
74
following functions. Note that the share package {\sf Polycyclic} must be
75
installed to use these functions.
76
77
\> AlmostCrystallographicPcpGroup( <dim>, <type>, <parameters> )
78
\> AlmostCrystallographicPcpDim3( <type>, <parameters> )
79
\> AlmostCrystallographicPcpDim4( <type>, <parameters> )
80
81
The input is the same as for the corresponding matrix group functions.
82
The output is a pcp group isomorphic to the corresponding matrix group.
83
An explicit isomorphism from an almost crystallographic matrix group
84
to the corresponding pcp group can be obtained by the following function.
85
86
\> IsomorphismPcpGroup( <G> )
87
88
We can use the polycyclic presentations of almost crystallographic
89
groups to exhibit structure information on these groups. For example,
90
we can determine their Fitting subgroup and ask group-theoretic
91
questions about this nilpotent group. The factor $G / Fit(G)$ of an
92
almost crystallographic group $G$ is called *holonomy group*. We
93
provide access to this factor of a pcp group via the following
94
functions. Let $G$ be an almost crystallographic pcp group.
95
96
\> HolonomyGroup( <G> )
97
\> NaturalHomomorphismOnHolonomyGroup( <G> )
98
99
The following example shows applications of these functions.
100
101
\beginexample
102
gap> G := AlmostCrystallographicPcpGroup( 4, 50, [ 1, -4, 1, 2 ] );
103
Pcp-group with orders [ 4, 0, 0, 0, 0 ]
104
gap> Cgs(G);
105
[ g1, g2, g3, g4, g5 ]
106
107
gap> F := FittingSubgroup( G );
108
Pcp-group with orders [ 0, 0, 0, 0 ]
109
gap> Centre(F);
110
Pcp-group with orders [ 0, 0 ]
111
gap> LowerCentralSeries(F);
112
[ Pcp-group with orders [ 0, 0, 0, 0 ], Pcp-group with orders [ 0 ],
113
Pcp-group with orders [ ] ]
114
gap> UpperCentralSeries(F);
115
[ Pcp-group with orders [ 0, 0, 0, 0 ], Pcp-group with orders [ 0, 0 ],
116
Pcp-group with orders [ ] ]
117
gap> MinimalGeneratingSet(F);
118
[ g2, g3, g4 ]
119
120
gap> H := HolonomyGroup( G );
121
Pcp-group with orders [ 4 ]
122
gap> hom := NaturalHomomorphismOnHolonomyGroup( G );
123
[ g1, g2, g3, g4, g5 ] -> [ g1, identity, identity, identity, identity ]
124
gap> U := Subgroup( H, [Pcp(H)[1]^2] );
125
Pcp-group with orders [ 2 ]
126
gap> PreImage( hom, U );
127
Pcp-group with orders [ 2, 0, 0, 0, 0 ]
128
\endexample
129
130
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
131
\Section{More about the type and the defining parameters}
132
133
Each group from this library knows that it is almost crystallographic and,
134
additionally, it knows its type and defining parameters.
135
136
\> AlmostCrystallographicInfo( <G> )
137
138
This attribute is set for groups from the library only. It is not possible
139
at current to determine the type and the defining parameters for an arbitrary
140
almost crystallographic groups which is not defined by the library access
141
functions.
142
143
\beginexample
144
gap> G := AlmostCrystallographicGroup( 4, 70, false );
145
<matrix group of size infinity with 5 generators>
146
gap> IsAlmostCrystallographic(G);
147
true
148
gap> AlmostCrystallographicInfo(G);
149
rec( dim := 4, type := 70, param := [ 1, -4, 1, 2, -3 ] )
150
\endexample
151
152
\beginexample
153
gap> G := AlmostCrystallographicPcpGroup( 4, 70, false );
154
Pcp-group with orders [ 6, 0, 0, 0, 0 ]
155
gap> IsAlmostCrystallographic(G);
156
true
157
gap> AlmostCrystallographicInfo(G);
158
rec( dim := 4, type := 70, param := [ -3, 2, 5, 1, 0 ] )
159
\endexample
160
161
We consider the types of almost crystallographic groups in more detail. The
162
almost crystallographic groups in dimensions 3 and 4 fall into three families
163
164
\beginlist
165
\item{(1)} 3-dimensional almost crystallographic groups.
166
\item{(2)} 4-dimensional almost crystallographic groups with a
167
Fitting subgroup of class 2.
168
\item{(3)} 4-dimensional almost crystallographic groups with a
169
Fitting subgroup of class 3.
170
\endlist
171
172
These families are split up further into subfamilies in \cite{KD} and to
173
each subfamily is assigned a type; that is, a string which is used to
174
identify the subfamily. As mentioned above, for the 3-dimensional almost
175
crystallographic groups the type is a string representing the numbers from
176
1 to 17, i.e. the available types are `"01"', `"02"', $\ldots$, `"17"'.
177
178
For the 4-dimensional almost crystallographic groups with a Fitting subgroup
179
of class 2 the type is a string of 3 or 4 characters. In general, a string of
180
3 characters representing
181
the number of the table entry in \cite{KD} is used. So possible types are
182
`"001"', `"002"', $\ldots$. The reader is warned however that not all
183
possible numbers are used, e.g.\ there are no groups of type `"016"'. Also,
184
the types do not appear in their natural order in \cite{KD}. Moreover, for
185
certain numbers there is more than one family of groups listed in \cite{KD}.
186
For example, the 3 families of groups corresponding to number 19 on pages
187
179-180 of \cite{KD} have types `"019"', `"019b"' and `"019c"' (the order is
188
the one given in \cite{KD}).
189
190
For the last category of groups, the 4-dimensional almost crystallographic
191
groups with a Fitting subgroup of class 3, the type is a string of 2 or 3
192
characters, where the first character is always the letter `"B"'. This `"B"'
193
is followed by the number of the table entry as found in \cite{KD},
194
eventually followed by a `"b"' or `"c"' as in the previous case.
195
196
For each type of almost crystallographic group contained in the library
197
there exists a function taking a parameter list as input and returning
198
the desired matrix or pcp group. These functions can be accessed
199
from \GAP\ using the lists `ACDim3Funcs', `ACDim4Funcs', `ACPcpDim3Funcs'
200
and `ACPcpDim4Funcs' which consist of the corresponding functions.
201
202
Although we include these direct access functions here for completeness,
203
we note that the user should in general use the higher-level functions
204
introduced above to obtain almost crystallographic groups from the
205
library. In particular, these low-level access functions return matrix
206
or pcp groups, but the almost crystallographic info flags will not be
207
attached to them.
208
209
\beginexample
210
gap> ACDim3Funcs[15];
211
function( k1, k2, k3, k4 ) ... end
212
gap> ACDim3Funcs[15](1,1,1,1);
213
<matrix group with 5 generators>
214
gap> ACPcpDim3Funcs[1](1);
215
Pcp-group with orders [ 0, 0, 0 ]
216
\endexample
217
218
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
219
\Section{The electronic versus the printed library}
220
221
The package `aclib' can be considered as the electronic version of
222
Chapter 7 of \cite{KD}. In this section we outline the relationship
223
between the library presented in this manual and the printed version
224
in \cite{KD}. First we consider an example. At page 175 of \cite{KD},
225
we find the following groups in the table starting with entry ``13''.
226
227
\medskip
228
13. $Q=P2/c$
229
$$
230
\matrix{ E:\;\langle a,b,c,d,\alpha,\beta\;\|\; & {\,[b,a]=1}\hskip 1.61cm
231
{[d,a]=1}
232
\hfill & \rangle \cr
233
& \matrix{ {[c,a]=d^{2 k_1}}\hfill & {[d,b]=1}\hfill \cr
234
{[c,b]=1}\hfill & {[d,c]=1}\hfill \cr
235
\alpha a=a^{-1}\alpha d^{k_{2}}\hfill & \alpha^2=d^{k_3}\hfill \cr
236
\alpha b=b\alpha \hfill & \alpha d= d \alpha\hfill \cr
237
\alpha c=c^{-1}\alpha d^{-2 k_6}\hfill & \cr
238
\beta a=a^{-1}\beta d^{k_1+k_2} \hfill & \beta^2=d^{k_5}\hfill \cr
239
\beta b=b^{-1}\beta d^{k_4}\hfill & \beta d= d \beta\hfill \cr
240
\beta c=c^{-1}\beta d^{-2 k_6}\hfill & \alpha \beta=c\beta\alpha d^{k_6}
241
\hfill } & }
242
$$
243
244
$$\lambda(\alpha)=\left(\matrix{
245
1& \frac{k_1}{2}+k_2 & 0 & -2 k_6 & \frac{k_3}{2}+\frac{k_6}{2} \cr
246
0& -1 & 0 & 0 & 0 \cr
247
0 & 0& 1 & 0 & 0 \cr
248
0 & 0 & 0 & -1 & \frac12\cr
249
0 & 0 & 0 & 0 & 1
250
}\right)
251
\;\;\lambda(\beta)=\left(\matrix{
252
1& k_1+k_2 & k_4 & -2 k_6 & \frac{k_5}{2} \cr
253
0& -1 & 0 & 0 & 0 \cr
254
0 & 0& -1 & 0 & 0 \cr
255
0 & 0 & 0 & -1 & 0\cr
256
0 & 0 & 0 & 0 & 1
257
}\right)
258
$$
259
260
$$H^2(Q,\Z{})=\Z{}\oplus(\Z{}_2)^4=\Z^{6}/A,$$
261
$$A=\{(k_1,\ldots,k_6)\|
262
k_1=0,\;k_2,\ldots, k_5\in 2\Z{},\;k_6\in\Z\}$$
263
AB-groups:
264
265
$\forall k>0,\;k\equiv 0\bmod 2,\;<(k,0,1,0,1,0)>$
266
267
\medskip
268
269
The number ``13'' at the beginning of this entry is the type of the
270
almost crystallographic group in this library. This family of groups
271
with type 13 depends on 6 parameters $k_1, k_2, \ldots, k_6$ and these
272
are the <parameters> list in this library. The rational matrix
273
representation in \GAP\ corresponds exactly to the printed version in
274
\cite{KD} where it is named $\lambda$. In the example below, we consider
275
the group with parameters $(k_1,k_2,k_3,k_4,k_5,k_6)=(8,0,1,0,1,0)$.
276
277
\beginexample
278
gap> G:=AlmostCrystallographicDim4("013",[8,0,1,0,1,0]);
279
<matrix group with 6 generators>
280
gap> G.5;
281
[ [ 1, 4, 0, 0, 1/2 ], [ 0, -1, 0, 0, 0 ], [ 0, 0, 1, 0, 0 ],
282
[ 0, 0, 0, -1, 1/2 ], [ 0, 0, 0, 0, 1 ] ]
283
gap> G.6;
284
[ [ 1, 8, 0, 0, 1/2 ], [ 0, -1, 0, 0, 0 ], [ 0, 0, -1, 0, 0 ],
285
[ 0, 0, 0, -1, 0 ], [ 0, 0, 0, 0, 1 ] ]
286
\endexample
287
288
For a 4-dimensional almost crystallographic group the matrix group is
289
built up such that $\{ a, b, c, d, \alpha, \beta, \gamma \}$ as described
290
in \cite{KD} forms the defining generating set of $G$. For certain types
291
the elements $\alpha$, $\beta$ or $\gamma$ may not be present.
292
Similarly, for a 3-dimensional group we have the generating set $\{ a, b,
293
c, \alpha, \beta \}$ and $\alpha$ and $\beta$ may be absent.
294
295
\bigskip
296
297
To obtain a polycyclic generating sequence from the defining generators
298
of the matrix group we have to order the elements in the generating set
299
suitably. For this purpose we take the subsequence of $(\gamma, \beta,
300
\alpha, a, b, c, d)$ of those generators which are present in the
301
defining generating set of the matrix group. This new ordering of the
302
generators is then used to define a polycyclic presentation of the given
303
almost crystallographic group.
304
305
306