14500
|
1 |
%
|
|
2 |
\begin{isabellebody}%
|
|
3 |
\def\isabellecontext{a{\isadigit{5}}}%
|
|
4 |
\isamarkupfalse%
|
|
5 |
%
|
|
6 |
\isamarkupsubsection{The Euclidean Algorithm -- Inductively%
|
|
7 |
}
|
|
8 |
\isamarkuptrue%
|
|
9 |
%
|
|
10 |
\isamarkupsubsubsection{Rules without Base Case%
|
|
11 |
}
|
|
12 |
\isamarkuptrue%
|
|
13 |
%
|
|
14 |
\begin{isamarkuptext}%
|
|
15 |
Show that the following%
|
|
16 |
\end{isamarkuptext}%
|
|
17 |
\isamarkuptrue%
|
|
18 |
\isacommand{consts}\ evenempty\ {\isacharcolon}{\isacharcolon}\ {\isachardoublequote}nat\ set{\isachardoublequote}\isanewline
|
|
19 |
\isamarkupfalse%
|
|
20 |
\isacommand{inductive}\ evenempty\isanewline
|
|
21 |
\ \ \isakeyword{intros}\isanewline
|
|
22 |
\ \ Add{\isadigit{2}}Ie{\isacharcolon}\ {\isachardoublequote}n\ {\isasymin}\ evenempty\ {\isasymLongrightarrow}\ Suc{\isacharparenleft}Suc\ n{\isacharparenright}\ {\isasymin}\ evenempty{\isachardoublequote}\isamarkupfalse%
|
|
23 |
%
|
|
24 |
\begin{isamarkuptext}%
|
|
25 |
defines the empty set:%
|
|
26 |
\end{isamarkuptext}%
|
|
27 |
\isamarkuptrue%
|
|
28 |
\isacommand{lemma}\ evenempty{\isacharunderscore}empty{\isacharcolon}\ {\isachardoublequote}evenempty\ {\isacharequal}\ {\isacharbraceleft}{\isacharbraceright}{\isachardoublequote}\isamarkupfalse%
|
|
29 |
\isamarkupfalse%
|
|
30 |
%
|
|
31 |
\isamarkupsubsubsection{The Euclidean Algorithm%
|
|
32 |
}
|
|
33 |
\isamarkuptrue%
|
|
34 |
%
|
|
35 |
\begin{isamarkuptext}%
|
|
36 |
Define inductively the set \isa{gcd}, which characterizes
|
|
37 |
the greatest common divisor of two natural numbers:%
|
|
38 |
\end{isamarkuptext}%
|
|
39 |
\isamarkuptrue%
|
|
40 |
\ \ gcd\ {\isacharcolon}{\isacharcolon}\ {\isachardoublequote}{\isacharparenleft}nat\ {\isasymtimes}\ nat\ {\isasymtimes}\ nat{\isacharparenright}\ set{\isachardoublequote}\isamarkupfalse%
|
|
41 |
%
|
|
42 |
\begin{isamarkuptext}%
|
|
43 |
Here, \isa{{\isacharparenleft}a{\isacharcomma}b{\isacharcomma}g{\isacharparenright}\ {\isasymin}\ gcd} means that \isa{g} is the gcd
|
|
44 |
of \isa{a} und \isa{b}. The definition should closely follow the
|
|
45 |
Euclidean algorithm.
|
|
46 |
|
|
47 |
Reminder: The Euclidean algorithm repeatedly subtracts the smaller
|
|
48 |
from the larger number, until one of the numbers is 0. Then, the other
|
|
49 |
number is the gcd.%
|
|
50 |
\end{isamarkuptext}%
|
|
51 |
\isamarkuptrue%
|
|
52 |
%
|
|
53 |
\begin{isamarkuptext}%
|
|
54 |
Now, compute the gcd of 15 and 10:%
|
|
55 |
\end{isamarkuptext}%
|
|
56 |
\isamarkuptrue%
|
|
57 |
\isacommand{lemma}\ {\isachardoublequote}{\isacharparenleft}{\isadigit{1}}{\isadigit{5}}{\isacharcomma}\ {\isadigit{1}}{\isadigit{0}}{\isacharcomma}\ {\isacharquery}g{\isacharparenright}\ \ {\isasymin}\ gcd{\isachardoublequote}\isamarkupfalse%
|
|
58 |
\isamarkupfalse%
|
|
59 |
%
|
|
60 |
\begin{isamarkuptext}%
|
|
61 |
How does your algorithm behave on special cases as the following?%
|
|
62 |
\end{isamarkuptext}%
|
|
63 |
\isamarkuptrue%
|
|
64 |
\isacommand{lemma}\ {\isachardoublequote}{\isacharparenleft}{\isadigit{0}}{\isacharcomma}\ {\isadigit{0}}{\isacharcomma}\ {\isacharquery}g{\isacharparenright}\ \ {\isasymin}\ gcd{\isachardoublequote}\isamarkupfalse%
|
|
65 |
\isamarkupfalse%
|
|
66 |
%
|
|
67 |
\begin{isamarkuptext}%
|
|
68 |
Show that the gcd is really a divisor (for the proof, you need an appropriate lemma):%
|
|
69 |
\end{isamarkuptext}%
|
|
70 |
\isamarkuptrue%
|
|
71 |
\isacommand{lemma}\ gcd{\isacharunderscore}divides{\isacharcolon}\ {\isachardoublequote}{\isacharparenleft}a{\isacharcomma}b{\isacharcomma}g{\isacharparenright}\ {\isasymin}\ gcd\ {\isasymLongrightarrow}\ g\ dvd\ a\ {\isasymand}\ g\ dvd\ b{\isachardoublequote}\isamarkupfalse%
|
|
72 |
\isamarkupfalse%
|
|
73 |
%
|
|
74 |
\begin{isamarkuptext}%
|
|
75 |
Show that the gcd is the greatest common divisor:%
|
|
76 |
\end{isamarkuptext}%
|
|
77 |
\isamarkuptrue%
|
|
78 |
\isacommand{lemma}\ gcd{\isacharunderscore}greatest\ {\isacharbrackleft}rule{\isacharunderscore}format{\isacharbrackright}{\isacharcolon}\ {\isachardoublequote}{\isacharparenleft}a{\isacharcomma}b{\isacharcomma}g{\isacharparenright}\ {\isasymin}\ gcd\ {\isasymLongrightarrow}\isanewline
|
|
79 |
\ \ {\isadigit{0}}\ {\isacharless}\ a\ {\isasymor}\ {\isadigit{0}}\ {\isacharless}\ b\ {\isasymlongrightarrow}\ {\isacharparenleft}{\isasymforall}\ d{\isachardot}\ d\ dvd\ a\ {\isasymlongrightarrow}\ d\ dvd\ b\ {\isasymlongrightarrow}\ d\ {\isasymle}\ g{\isacharparenright}{\isachardoublequote}\isamarkupfalse%
|
|
80 |
\isamarkupfalse%
|
|
81 |
%
|
|
82 |
\begin{isamarkuptext}%
|
|
83 |
Here as well, you will have to prove a suitable lemma. What is
|
|
84 |
the precondition \isa{{\isadigit{0}}\ {\isacharless}\ a\ {\isasymor}\ {\isadigit{0}}\ {\isacharless}\ b} good for?
|
|
85 |
|
|
86 |
So far, we have only shown that \isa{gcd} is correct, but your
|
|
87 |
algorithm might not compute a result for all values \isa{a{\isacharcomma}b}. Thus, show completeness of the algorithm:%
|
|
88 |
\end{isamarkuptext}%
|
|
89 |
\isamarkuptrue%
|
|
90 |
\isacommand{lemma}\ gcd{\isacharunderscore}defined{\isacharcolon}\ {\isachardoublequote}{\isasymforall}\ a\ b{\isachardot}\ {\isasymexists}\ g{\isachardot}\ {\isacharparenleft}a{\isacharcomma}\ b{\isacharcomma}\ g{\isacharparenright}\ {\isasymin}\ gcd{\isachardoublequote}\isamarkupfalse%
|
|
91 |
\isamarkupfalse%
|
|
92 |
%
|
|
93 |
\begin{isamarkuptext}%
|
|
94 |
The following lemma, proved by course-of-value recursion over
|
|
95 |
\isa{n}, may be useful. Why does standard induction over natural
|
|
96 |
numbers not work here?%
|
|
97 |
\end{isamarkuptext}%
|
|
98 |
\isamarkuptrue%
|
|
99 |
\isacommand{lemma}\ gcd{\isacharunderscore}defined{\isacharunderscore}aux\ {\isacharbrackleft}rule{\isacharunderscore}format{\isacharbrackright}{\isacharcolon}\ \isanewline
|
|
100 |
\ \ {\isachardoublequote}{\isasymforall}\ a\ b{\isachardot}\ {\isacharparenleft}a\ {\isacharplus}\ b{\isacharparenright}\ {\isasymle}\ n\ {\isasymlongrightarrow}\ {\isacharparenleft}{\isasymexists}\ g{\isachardot}\ {\isacharparenleft}a{\isacharcomma}\ b{\isacharcomma}\ g{\isacharparenright}\ {\isasymin}\ gcd{\isacharparenright}{\isachardoublequote}\isanewline
|
|
101 |
\isamarkupfalse%
|
|
102 |
\isacommand{apply}\ {\isacharparenleft}induct\ rule{\isacharcolon}\ nat{\isacharunderscore}less{\isacharunderscore}induct{\isacharparenright}\isanewline
|
|
103 |
\isamarkupfalse%
|
|
104 |
\isacommand{apply}\ clarify\isanewline
|
|
105 |
\isamarkupfalse%
|
|
106 |
\isamarkupfalse%
|
|
107 |
%
|
|
108 |
\begin{isamarkuptext}%
|
|
109 |
The idea is to show that \isa{gcd} yields a result for all
|
|
110 |
\isa{a{\isacharcomma}\ b} whenever it is known that \isa{gcd} yields a result
|
|
111 |
for all \isa{a{\isacharprime}{\isacharcomma}\ b{\isacharprime}} whose sum is smaller than \isa{a\ {\isacharplus}\ b}.
|
|
112 |
|
|
113 |
In order to prove this lemma, make case distinctions corresponding to
|
|
114 |
the different clauses of the algorithm, and show how to reduce
|
|
115 |
computation of \isa{gcd} for \isa{a{\isacharcomma}\ b} to computation of \isa{gcd} for suitable smaller \isa{a{\isacharprime}{\isacharcomma}\ b{\isacharprime}}.%
|
|
116 |
\end{isamarkuptext}%
|
|
117 |
\isamarkuptrue%
|
|
118 |
\isanewline
|
|
119 |
\isamarkupfalse%
|
|
120 |
\end{isabellebody}%
|
|
121 |
%%% Local Variables:
|
|
122 |
%%% mode: latex
|
|
123 |
%%% TeX-master: "root"
|
|
124 |
%%% End:
|