Datasets:

Modalities:
Text
Languages:
English
Libraries:
Datasets
License:
File size: 10,535 Bytes
4365a98
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
(*
  File:   Master_Theorem_Examples.thy
  Author: Manuel Eberl <manuel@pruvisto.org>

  Examples for the application of the Master theorem and related proof methods.
*)

section \<open>Examples\<close>
theory Master_Theorem_Examples
imports
  Complex_Main
  Akra_Bazzi_Method
  Akra_Bazzi_Approximation
begin

subsection \<open>Merge sort\<close>

(* A merge sort cost function that is parametrised with the recombination costs *)
function merge_sort_cost :: "(nat \<Rightarrow> real) \<Rightarrow> nat \<Rightarrow> real" where
  "merge_sort_cost t 0 = 0"
| "merge_sort_cost t 1 = 1"
| "n \<ge> 2 \<Longrightarrow> merge_sort_cost t n = 
     merge_sort_cost t (nat \<lfloor>real n / 2\<rfloor>) + merge_sort_cost t (nat \<lceil>real n / 2\<rceil>) + t n"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma merge_sort_nonneg[simp]: "(\<And>n. t n \<ge> 0) \<Longrightarrow> merge_sort_cost t x \<ge> 0"
  by (induction t x rule: merge_sort_cost.induct) (simp_all del: One_nat_def)

lemma "t \<in> \<Theta>(\<lambda>n. real n) \<Longrightarrow> (\<And>n. t n \<ge> 0) \<Longrightarrow> merge_sort_cost t \<in> \<Theta>(\<lambda>n. real n * ln (real n))"
  by (master_theorem 2.3) simp_all

subsection \<open>Karatsuba multiplication\<close>

function karatsuba_cost :: "nat \<Rightarrow> real" where
  "karatsuba_cost 0 = 0"
| "karatsuba_cost 1 = 1"
| "n \<ge> 2 \<Longrightarrow> karatsuba_cost n = 
     3 * karatsuba_cost (nat \<lceil>real n / 2\<rceil>) + real n"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma karatsuba_cost_nonneg[simp]: "karatsuba_cost n \<ge> 0"
  by (induction n rule: karatsuba_cost.induct) (simp_all del: One_nat_def)

lemma "karatsuba_cost \<in> O(\<lambda>n. real n powr log 2 3)"
   by (master_theorem 1 p': 1) (simp_all add: powr_divide)

lemma karatsuba_cost_pos: "n \<ge> 1 \<Longrightarrow> karatsuba_cost n > 0"
  by (induction n rule: karatsuba_cost.induct) (auto intro!: add_nonneg_pos simp del: One_nat_def)

lemma "karatsuba_cost \<in> \<Theta>(\<lambda>n. real n powr log 2 3)"
  using karatsuba_cost_pos
  by (master_theorem 1 p': 1) (auto simp add: powr_divide eventually_at_top_linorder)


subsection \<open>Strassen matrix multiplication\<close>

function strassen_cost :: "nat \<Rightarrow> real" where
  "strassen_cost 0 = 0"
| "strassen_cost 1 = 1"
| "n \<ge> 2 \<Longrightarrow> strassen_cost n = 7 * strassen_cost (nat \<lceil>real n / 2\<rceil>) + real (n^2)"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma strassen_cost_nonneg[simp]: "strassen_cost n \<ge> 0"
  by (induction n rule: strassen_cost.induct) (simp_all del: One_nat_def)

lemma "strassen_cost \<in> O(\<lambda>n. real n powr log 2 7)"
  by (master_theorem 1 p': 2) (auto simp: powr_divide eventually_at_top_linorder)

lemma strassen_cost_pos: "n \<ge> 1 \<Longrightarrow> strassen_cost n > 0"
  by (cases n rule: strassen_cost.cases) (simp_all add: add_nonneg_pos del: One_nat_def)

lemma "strassen_cost \<in> \<Theta>(\<lambda>n. real n powr log 2 7)"
  using strassen_cost_pos
  by (master_theorem 1 p': 2) (auto simp: powr_divide eventually_at_top_linorder)


subsection \<open>Deterministic select\<close>

(* This is not possible with the standard Master theorem from literature *)
function select_cost :: "nat \<Rightarrow> real" where
  "n \<le> 20 \<Longrightarrow> select_cost n = 0"
| "n > 20 \<Longrightarrow> select_cost n = 
     select_cost (nat \<lfloor>real n / 5\<rfloor>) + select_cost (nat \<lfloor>7 * real n / 10\<rfloor> + 6) + 12 * real n / 5"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma "select_cost \<in> \<Theta>(\<lambda>n. real n)"
  by (master_theorem 3) auto


subsection \<open>Decreasing function\<close>

function dec_cost :: "nat \<Rightarrow> real" where
  "n \<le> 2 \<Longrightarrow> dec_cost n = 1"
| "n > 2 \<Longrightarrow> dec_cost n = 0.5*dec_cost (nat \<lfloor>real n / 2\<rfloor>) + 1 / real n"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma "dec_cost \<in> \<Theta>(\<lambda>x::nat. ln x / x)"
  by (master_theorem 2.3) simp_all


subsection \<open>Example taken from Drmota and Szpakowski\<close>

function drmota1 :: "nat \<Rightarrow> real" where
  "n < 20 \<Longrightarrow> drmota1 n = 1"
| "n \<ge> 20 \<Longrightarrow> drmota1 n = 2 * drmota1 (nat \<lfloor>real n/2\<rfloor>) + 8/9 * drmota1 (nat \<lfloor>3*real n/4\<rfloor>) + real n^2 / ln (real n)"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma "drmota1 \<in> \<Theta>(\<lambda>n::real. n^2 * ln (ln n))"
  by (master_theorem 2.2) (simp_all add: power_divide)


function drmota2 :: "nat \<Rightarrow> real" where
  "n < 20 \<Longrightarrow> drmota2 n = 1"
| "n \<ge> 20 \<Longrightarrow> drmota2 n = 1/3 * drmota2 (nat \<lfloor>real n/3 + 1/2\<rfloor>) + 2/3 * drmota2 (nat \<lfloor>2*real n/3 - 1/2\<rfloor>) + 1"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma "drmota2 \<in> \<Theta>(\<lambda>x. ln (real x))"
  by master_theorem simp_all

(* Average phrase length of Boncelet arithmetic coding. See Drmota and Szpankowski. *)
lemma boncelet_phrase_length:
  fixes p \<delta> :: real assumes p: "p > 0" "p < 1" and \<delta>: "\<delta> > 0" "\<delta> < 1" "2*p + \<delta> < 2"
  fixes d :: "nat \<Rightarrow> real"
  defines "q \<equiv> 1 - p"
  assumes d_nonneg: "\<And>n. d n \<ge> 0"
  assumes d_rec: "\<And>n. n \<ge> 2 \<Longrightarrow> d n = 1 + p * d (nat \<lfloor>p * real n + \<delta>\<rfloor>) + q * d (nat \<lfloor>q * real n - \<delta>\<rfloor>)"
  shows   "d \<in> \<Theta>(\<lambda>x. ln x)"
  using assms by (master_theorem recursion: d_rec, simp_all)



subsection \<open>Transcendental exponents\<close>

(* Certain number-theoretic conjectures would imply that if all the parameters are rational,
   the Akra-Bazzi parameter is either rational or transcendental. That makes this case 
   probably transcendental *)
function foo_cost :: "nat \<Rightarrow> real" where
  "n < 200 \<Longrightarrow> foo_cost n = 0"
| "n \<ge> 200 \<Longrightarrow> foo_cost n = 
     foo_cost (nat \<lfloor>real n / 3\<rfloor>) + foo_cost (nat \<lfloor>3 * real n / 4\<rfloor> + 42) + real n"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma foo_cost_nonneg [simp]: "foo_cost n \<ge> 0"
  by (induction n rule: foo_cost.induct) simp_all

lemma "foo_cost \<in> \<Theta>(\<lambda>n. real n powr akra_bazzi_exponent [1,1] [1/3,3/4])"
proof (master_theorem 1 p': 1) 
  have "\<forall>n\<ge>200. foo_cost n > 0" by (simp add: add_nonneg_pos)
  thus "eventually (\<lambda>n. foo_cost n > 0) at_top" unfolding eventually_at_top_linorder by blast
qed simp_all

lemma "akra_bazzi_exponent [1,1] [1/3,3/4] \<in> {1.1519623..1.1519624}"
  by (akra_bazzi_approximate 29)


subsection \<open>Functions in locale contexts\<close>

locale det_select =
  fixes b :: real
  assumes b: "b > 0" "b < 7/10"
begin

function select_cost' :: "nat \<Rightarrow> real" where
  "n \<le> 20 \<Longrightarrow> select_cost' n = 0"
| "n > 20 \<Longrightarrow> select_cost' n = 
     select_cost' (nat \<lfloor>real n / 5\<rfloor>) + select_cost' (nat \<lfloor>b * real n\<rfloor> + 6) + 6 * real n + 5"
by force simp_all
termination using b by akra_bazzi_termination simp_all

lemma "a \<ge> 0 \<Longrightarrow> select_cost' \<in> \<Theta>(\<lambda>n. real n)"
  using b by (master_theorem 3, force+)

end


subsection \<open>Non-curried functions\<close>

(* Note: either a or b could be seen as recursion variables. *)
function baz_cost :: "nat \<times> nat \<Rightarrow> real" where
  "n \<le> 2 \<Longrightarrow> baz_cost (a, n) = 0"
| "n > 2 \<Longrightarrow> baz_cost (a, n) = 3 * baz_cost (a, nat \<lfloor>real n / 2\<rfloor>) + real a"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma baz_cost_nonneg [simp]: "a \<ge> 0 \<Longrightarrow> baz_cost (a, n) \<ge> 0"
  by (induction a n rule: baz_cost.induct[split_format (complete)]) simp_all

lemma
  assumes "a > 0"
  shows   "(\<lambda>x. baz_cost (a, x)) \<in> \<Theta>(\<lambda>x. x powr log 2 3)"
proof (master_theorem 1 p': 0)
  from assms have "\<forall>x\<ge>3. baz_cost (a, x) > 0" by (auto intro: add_nonneg_pos)
  thus "eventually (\<lambda>x. baz_cost (a, x) > 0) at_top" by (force simp: eventually_at_top_linorder)
qed (insert assms, simp_all add: powr_divide)

(* Non-"Akra-Bazzi" variables may even be modified without impacting the termination proof.
   However, the Akra-Bazzi theorem and the Master theorem itself do not apply anymore, 
   because bar_cost cannot be seen as a recursive function with one parameter *)
function bar_cost :: "nat \<times> nat \<Rightarrow> real" where
  "n \<le> 2 \<Longrightarrow> bar_cost (a, n) = 0"
| "n > 2 \<Longrightarrow> bar_cost (a, n) = 3 * bar_cost (2 * a, nat \<lfloor>real n / 2\<rfloor>) + real a"
by force simp_all
termination by akra_bazzi_termination simp_all


subsection \<open>Ham-sandwich trees\<close>
(* f(n) = f(n/4) + f(n/2) + 1 *)
function ham_sandwich_cost :: "nat \<Rightarrow> real" where
  "n < 4 \<Longrightarrow> ham_sandwich_cost n = 1"
| "n \<ge> 4 \<Longrightarrow> ham_sandwich_cost n = 
      ham_sandwich_cost (nat \<lfloor>n/4\<rfloor>) + ham_sandwich_cost (nat \<lfloor>n/2\<rfloor>) + 1"
by force simp_all
termination by akra_bazzi_termination simp_all

lemma ham_sandwich_cost_pos [simp]: "ham_sandwich_cost n > 0"
  by (induction n rule: ham_sandwich_cost.induct) simp_all


text \<open>The golden ratio\<close>

definition "\<phi> = ((1 + sqrt 5) / 2 :: real)"

lemma \<phi>_pos [simp]: "\<phi> > 0" and \<phi>_nonneg [simp]: "\<phi> \<ge> 0" and \<phi>_nonzero [simp]: "\<phi> \<noteq> 0"
proof-
  show "\<phi> > 0" unfolding \<phi>_def by (simp add: add_pos_nonneg)
  thus "\<phi> \<ge> 0" "\<phi> \<noteq> 0" by simp_all
qed


lemma "ham_sandwich_cost \<in> \<Theta>(\<lambda>n. n powr (log 2 \<phi>))"
proof (master_theorem 1 p': 0)
  have "(1 / 4) powr log 2 \<phi> + (1 / 2) powr log 2 \<phi> =
            inverse (2 powr log 2 \<phi>)^2 + inverse (2 powr log 2 \<phi>)"
        by (simp add: powr_divide field_simps powr_powr power2_eq_square powr_mult[symmetric]
                 del: powr_log_cancel)
  also have "... = inverse (\<phi>^2) + inverse \<phi>" by (simp add: power2_eq_square)
  also have "\<phi> + 1 = \<phi>*\<phi>" by (simp add: \<phi>_def field_simps)
  hence "inverse (\<phi>^2) + inverse \<phi> = 1" by (simp add: field_simps power2_eq_square)
  finally show "(1 / 4) powr log 2 \<phi> + (1 / 2) powr log 2 \<phi> = 1" by simp
qed simp_all

end