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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446
|
Require Import
List NPeano
QArith Qabs Qpossec Qsums Qround
Qmetric ZArith
CRArith CRsum (*AbstractIntegration*)
util.Qgcd
Program
uneven_CRplus
stdlib_omissions.P
stdlib_omissions.Z
stdlib_omissions.Q.
Open Scope uc_scope.
Set Automatic Introduction.
Hint Resolve Qpos_nonzero.
Hint Immediate Q.Qle_nat.
Hint Resolve Qmult_le_0_compat.
Hint Resolve QnonNeg.Qplus_nonneg.
Parameter (z:Z).
(* Zsqrt_plain_is_pos *)
(*
Lemma Zsqrt_r_nonneg (z: Z) (E: 0 <= z): (0 <= Zsqrt z)%Z.
Proof with auto.
destruct Zsqrt; try easy.
admit.
subst. simpl. omega.
Qed.
*)
Require Import Coq.ZArith.Zsqrt_compat.
Open Scope Z_scope.
Definition Z_4th_root_floor (x: Z): (0 <= x)%Z ->
{s: Z & {r: Z | x = Zpower s 4 + r /\ Zpower s 4 <= x < Zpower (s + 1) 4}}%Z.
Proof.
intro E.
destruct x.
exists 0%Z.
exists 0%Z.
split. reflexivity.
change (0 <= 0 < 1).
omega.
exists (projT1 (Zsqrt (projT1 (Zsqrt p (Zle_0_pos p))) (Zsqrt_plain_is_pos p E))).
set (Zsqrt (projT1 (Zsqrt p (Zle_0_pos p))) (Zsqrt_plain_is_pos p E)).
admit.
exfalso. apply E. reflexivity.
Defined.
Definition Z_4th_root_floor_plain (z: Z): Z :=
match z with
| Zpos p => projT1 (Z_4th_root_floor p (Zle_0_pos p))
| _ => 0%Z
end.
Lemma Zle_uniq {x y: Z} (p q: Zle x y): p = q.
Admitted.
Goal forall z, Z_4th_root_floor_plain z = Zsqrt_plain (Zsqrt_plain z).
Proof.
intros.
unfold Zsqrt_plain.
destruct z; try reflexivity.
unfold Z_4th_root_floor_plain, Z_4th_root_floor.
unfold projT1 at 1.
generalize (Zsqrt_plain_is_pos). (* p (Zle_0_pos p)). *)
unfold Zsqrt_plain.
(*generalize (Zsqrt).*) (*p (Zle_0_pos p)).*)
admit.
(*
destruct s.
simpl @projT1 at 1.
destruct x.
simpl. reflexivity. *)
admit.
admit.
(*
rewrite (Zle_uniq z (Zle_0_pos p)).
intro.
reflexivity.
simpl. intro. exfalso. apply z. reflexivity.*)
Qed.
Definition Q_4th_root_floor_plain (q: Q): Z := Z_4th_root_floor_plain (Qceiling q).
Section definition.
Context
(f: Q_as_MetricSpace --> CR)
(b: Q). (* bound for the absolute value of f's fourth derivative *)
Section approx.
Context (n : positive)(fr: Q) (w: Qpos) (e: Qpos).
Definition N: positive := P_of_succ_nat (Zabs_nat (Q_4th_root_floor_plain ((w^5) / 2880 * b / e))).
(* This Zabs is silly because we know the squaring thing only returns nonnegatives, but whatever. *)
(* Also, a ceil variant would obviate need to take the successor, but I haven't defined ceil variants
of the 4th root for Z/Q yet. *)
Definition iw : Qpos := (w / N)%Qpos.
Definition iw1 : Qpos := (w / n)%Qpos.
Definition halfiw : Qpos := (w / ((2#1) * N))%Qpos.
Definition halfiw1 : Qpos := (w / ((2#1) * n))%Qpos.
Open Scope Q_scope.
Definition simpson (fr: Q): CR :=
(' (iw / 6) * (f fr + f (fr + halfiw)%Q * '4 + f (fr + iw)%Q))%CR.
Definition simpson1 (fr: Q): CR :=
(' (iw1) * (f fr + f (fr + halfiw1)%Q * '4 + f (fr + iw1)%Q))%CR.
Definition approx: CR := CRsum_list (map (fun i: nat => simpson (fr + i * iw)) (N.enum (nat_of_P N))).
Definition approx1 : CR :=
CRsum_list (map (fun i: nat => simpson1 (fr + i * iw1)) (N.enum (nat_of_P n))).
End approx.
Lemma regular fr w: is_RegularFunction_noInf CR (approx fr w).
Admitted.
Definition simpson_integral fr w: CR := Cjoin (mkRegularFunction ('(0%Q))%CR (regular fr w)).
(*
Global Instance integrate: Integral f := @integral_extended_to_nn_width f pre_result.
*)
End definition.
Require Import ARtrans.
Require Import Qdlog.
Require Import BigQ ARbigQ ARQ ARbigD.
Definition eps (n : positive) := (1 # (10^n))%Qpos.
Definition answer (n:positive) (r:CR) : Z :=
let m := (10^n)%positive in
let (a,b) := ((approximate r (1#m)%Qpos) * m)%Q in
Zdiv a b.
(*Time Eval vm_compute in approximate (simpson_integral sin_uc 1 0 1) (1#100000)%Qpos.*)
Definition sum_pos `{Zero A, Plus A} (f : positive -> A) (n : positive) :=
Pos.peano_rect (λ _, A) 0 (λ p x, f p + x) n.
Definition sum_pos_iter `{Zero A, Plus A} (f : positive -> A) (n : positive) : A :=
match n with
| xH => 0
| _ =>
let z :=
Pos.iter
(Pos.pred n)
(λ y : positive * A, let (p, x) := y in ((Pos.succ p), (f p + x)))
(1%positive, 0) in
snd z
end.
Section ARsum.
Context `{AppRationals AQ}.
Definition ARsum_list_raw (l : list AR) (e : QposInf) : AQ :=
fold_left (@plus AQ _)
match l with
| nil => nil
| cons h t =>
let e' := QposInf_mult (1#(Pos.of_nat (length t)))%Qpos e in
(map (fun x => approximate x e') l)
end
0.
Definition ARsum_raw (f : positive -> AR) (n : positive) (eps : QposInf) : AQ :=
let e := (eps * (1 # Pos.pred n)%Qpos)%QposInf in
sum_pos_iter (λ p, approximate (f p) e) n.
Lemma ARsum_list_prf : forall l, @is_RegularFunction AQ_as_MetricSpace (ARsum_list_raw l).
Admitted.
Lemma ARsum_prf : forall f n, @is_RegularFunction AQ_as_MetricSpace (ARsum_raw f n).
Admitted.
Definition ARsum_list (l : list AR) : AR := Build_RegularFunction (ARsum_list_prf l).
Definition ARsum (f : positive -> AR) (n : positive) : AR := Build_RegularFunction (ARsum_prf f n).
End ARsum.
Section ARInt.
Context
`{AppRationals AQ}
(f : AR -> AR)
(B : Q) (* bound for the absolute value of f's fourth derivative *)
(a b : AR) (w : AQ).
Let width : AR := b - a.
Section ARIntN.
Variable n : positive.
Section ARIntEps.
Variable eps : Qpos.
Let hl' : AR := width * AQinv ('(Zpos n~0)). (* hl' = width / (2 * n) *)
Let eps' : Qpos := eps * (1 # (6 * n)%positive)%Qpos.
Let h (p : positive) := approximate (f (a + ARscale ('(Zpos p)) hl')) eps'.
Definition ARsimpson_sum_raw : AQ :=
4 * (sum_pos_iter (λ p, h (Pos.pred_double p)) (Pos.succ n)) +
2 * (sum_pos_iter (λ p, h p~0) n) +
(approximate (f a) eps' + approximate (f b) eps').
End ARIntEps.
Lemma ARsimson_sum_regular : is_RegularFunction_noInf AQ_as_MetricSpace ARsimpson_sum_raw.
Admitted.
Definition ARsimpson_sum : AR := mkRegularFunction 0 ARsimson_sum_regular.
End ARIntN.
Section ARIntEps1.
Variable eps : Qpos.
Definition num_intervals : nat := S (Z.to_nat (Q_4th_root_floor_plain ('w^5 / 2880 * B / eps))).
(* To be optimized *)
Definition num_intervals1 : positive :=
P_of_succ_nat (Zabs_nat (Q_4th_root_floor_plain (('w^5) / 2880 * B / eps))).
Definition num_intervals2 : positive :=
let w : Q := 'approximate width (1#1000)%Qpos + (1#1000) in
Pos.succ (Z.to_pos (Q_4th_root_floor_plain (w^5 / 2880 * B / eps))).
(* half-length *)
Let hl : AR := width * AQinv ('(Zpos (num_intervals2~0)%positive)).
Let f' (n : nat) := f(a + '(n : Z) * 'w * AQinv ('(2 * (num_intervals : Z))%Z)).
Let g (p : positive) := f(a + ARscale ('(Zpos p)) hl).
(*Let h (p : positive) (e : Qpos) := approximate (f (a + ARscale ('(Zpos p)) hl)) e.*)
Definition ARsimpson_raw : AR :=
(ARscale 4 (ARsum_list (map (fun i : nat => f' (2 * i + 1)) (N.enum (num_intervals - 0)))) +
ARscale 2 (ARsum_list (map (fun i : nat => f' (2 * i + 2)) (N.enum (num_intervals - 1)))) +
(f' 0 + f' (2 * num_intervals))) * 'w * AQinv ('(6 * (num_intervals : Z))%Z).
Definition ARsimpson1_raw : AR :=
((ARscale 4 (ARsum (λ p, g (Pos.pred_double p)) (Pos.succ num_intervals2))) +
(ARscale 2 (ARsum (λ p, g p~0) num_intervals2)) +
(f a + f b))
* width * AQinv ('(6 * (num_intervals2 : Z))%Z).
(*Definition ARsimpson_sum_raw : AQ :=
let e := eps * (1 # (6 * num_intervals2)%positive)%Qpos in
4 * (sum_pos_iter (λ p, h (Pos.pred_double p) e) (Pos.succ num_intervals2)) +
2 * (sum_pos_iter (λ p, h p~0 e) num_intervals2) +
(approximate (f a) e + approximate (f b) e).*)
Definition ARsimpson2_raw : AR :=
ARsimpson_sum num_intervals2 * (width * AQinv ('Zpos (6 * num_intervals2)%positive)).
End ARIntEps1.
Lemma ARsimson_regular : is_RegularFunction_noInf AR ARsimpson_raw.
Admitted.
Lemma ARsimson1_regular : is_RegularFunction_noInf AR ARsimpson1_raw.
Admitted.
Lemma ARsimson2_regular : is_RegularFunction_noInf AR ARsimpson2_raw.
Admitted.
Definition ARsimpson : AR := Cjoin (mkRegularFunction 0 ARsimson_regular).
Definition ARsimpson1 : AR := Cjoin (mkRegularFunction 0 ARsimson1_regular).
Definition ARsimpson2 : AR := Cjoin (mkRegularFunction 0 ARsimson2_regular).
End ARInt.
(*Time Compute approximate (ARexp (AQ := bigD) 4) (eps 2000)
Time Check approximate ((ARexp (AQ := bigD) 4) * '((10 ^ 1000)%positive : Z)) (1#1)%Qpos.*)
(*Compute N 3 1 (eps 20).
Compute num_intervals (AQ := bigD) 3 1 (eps 13).*)
(*Extraction "mult.ml" ARmult.*)
(*Time Compute approximate (simpson_integral (exp_bound_uc 2) 3 0 1) (eps 11).*)
(*
(* The following shows that in evaluating x * y up to eps, (approximate x
(eps / (2 * c))) where c is an approximation of y up to 1, is computed once
and not twice. We make y very large so that the approximation of x takes a
long time. Multiplcation takes less than twice the time of the approximation of x. *)
Definition int := (ARsimpson (AQ := bigD) (ARexp_bounded (AQ := bigD) 2) 3 0 1).
Definition e := '((10 ^ 12)%positive : Z) : ARbigD.
Time Compute approximate (int * e) (1#1)%Qpos.
Time Compute approximate int (eps 13).
*)
(* (ARexp x) calls ARexp_bounded on (Qceiling ('approximate x (1#1)%Qpos + (1#1))) and x.
If x = 1, then the approximation is 2. *)
Definition repeat {A : Type} (M : unit -> A) (n : positive) :=
Pos.iter n (fun _ => (fun _ => tt) (M tt)) tt.
(*Definition M :=
fun _ : unit =>
approximate (ARexp_bounded (AQ := bigD) 2 1) (eps 12).*)
(*Compute num_intervals2 (AQ := bigD) 3 0 1 (eps 15).*)
(*Time Compute approximate (ARsimpson (AQ := bigD) (ARexp_bounded (AQ := bigD) 2) 3 0 1) (eps 14).
Time Compute approximate (ARsimpson1 (AQ := bigD) (ARexp_bounded (AQ := bigD) 2) 3 0 1) (eps 14).*)
(*Time Compute approximate (ARsimpson2 (AQ := Q) (ARexp_bounded 2) 3 0 1) (eps 9).
Time Compute approximate (ARsimpson_sum (AQ := bigD) (ARexp_bounded (AQ := bigD) 2) 0 1 1012) (eps 14).*)
Section Picard.
Context `{AppRationals AQ} (F : AR -> AR) (a b : AR).
Definition picard (f : AR -> AR) (x : AR) := b + ARsimpson2 (AQ := AQ) (λ t, F (f t)) 1 a x.
Definition picard_iter (n : nat) : AR -> AR := nat_iter n picard (λ _, b).
End Picard.
Definition d := approximate (picard_iter (AQ := bigD) (λ y, y) 0 1 6 1) (eps 1).
Extraction "simpson.ml" d.
Time Compute approximate (picard_iter (AQ := bigD) (λ y, y) 0 1 6 1) (eps 1).
(*Time Compute approximate (ARsimpson (AQ := bigD) ARexp 3 0 1) (eps 10).
Time Compute approximate (ARsimpson (AQ := bigD) ARarctan 1 0 1) (eps 1).
Time Compute approximate (ARsimpson (AQ := bigD) ARsqrt 3 0 1) (eps 12).
Timeout 30 Compute approximate (ARsimpson (AQ := bigD) ARexp 3 0 1) (eps 1).
Compute num_intervals (AQ := bigD) 3 1 (eps 0).*)
Section ARInt'.
Context
`{AppRationals AQ}
(f : AQ -> AR)
(B : Q). (* bound for the absolute value of f's fourth derivative *)
Section ARapprox.
Context (n : positive) (a : AQ) (w : AQ) (eps : Qpos).
Definition N' : nat := Z.to_nat (1 + Zdiv (Qdlog2 ('w^5 / 2880 * B / eps))%Q 4).
Definition iw' : AQ := w ≪ -(N' : Z).
Definition iw1' : AQ := w ≪ -(n : Z).
Definition simpson' (a' : AQ) : AR :=
('iw' * (f a' + f (a' + (iw' ≪ -1)) * '4 + f (a' + iw'))).
Definition simpson1' (a' : AQ) : AR :=
('iw1' * (f a' + f (a' + (iw1' ≪ -1)) * '4 + f (a' + iw1'))).
Definition approx' : AR :=
ARsum_list (map (fun i : nat => simpson' (a + '(i : Z) * iw')) (N.enum (2^N'))).
Definition approx1' : AR :=
ARsum_list (map (fun i : nat => simpson1' (a + '(i : Z) * iw1')) (N.enum (nat_of_P (2^n)%positive))).
End ARapprox.
Lemma regular' a w : is_RegularFunction_noInf AR (approx' a w).
Admitted.
Definition simpson_integral' a w : AR := Cjoin (mkRegularFunction 0 (regular' a w)).
End ARInt'.
Time Compute approximate (simpson_integral' (AQ := bigD) AQexp 3 0 1) (eps 10).
Time Compute approximate (simpson_integral' (AQ := bigD) ARexp 3 0 1) (eps 10).
(*Eval compute in N' (AQ := bigD) 1 1 (eps 8).
Eval compute in N 1 1 (eps 8).*)
(*Time Check approximate (ARexp_bounded_uc (AQ := bigD) 2 1) (eps 20).
Time Check approximate (ARexp (AQ := bigD) 1) (eps 20).
Time Eval vm_compute in approximate (ARexp_bounded_uc (AQ := bigD) 2 1) (eps 20).
Time Eval vm_compute in approximate (ARexp (AQ := bigD) 1) (eps 20).*)
(*Time Check approximate (Cjoin_fun (Cmap_fun AQPrelengthSpace (ARexp_bounded_uc (AQ := bigD) 2) 1)) (eps 20).
Time Eval vm_compute in
approximate (Cjoin_fun (Cmap_fun AQPrelengthSpace (ARexp_bounded_uc (AQ := bigD) 2) 1)) (eps 20).*)
Time Eval vm_compute in approximate (ARexp (AQ := bigD) 1) (eps 20).
Time Eval vm_compute in approximate (exp 1) (eps 20).
Time Eval vm_compute in approximate (exp_bound_uc 3 1) (eps 130).
Time Eval vm_compute in approximate (ARsin_uc (AQ := bigD) 1) (eps 20).
Time Eval vm_compute in approximate (sin_uc 1) (eps 20).
Time Eval vm_compute in approximate (sin_slow 1) (eps 50).
Time Eval vm_compute in approximate (ARsin (AQ := bigD) 1) (eps 50).
Require Import PowerSeries.
Time Eval vm_compute in
approximate (ARsin (AQ := bigD) (ARsin (AQ := bigD) (ARsin (AQ := bigD) 1))) (eps 25).
Time Eval vm_compute in approximate (approx1 sin_uc 32 0 1) (eps 50).
Time Eval vm_compute in approximate (approx1' (AQ := bigD) ARsin_uc 5 0 1) (eps 50).
Time Eval vm_compute in
(fun _ => tt) (map (fun _ => approximate (ARsin_uc (AQ := bigD) 1) (eps 10)) (N.enum 10)).
Time Eval vm_compute in
(fun _ => tt) (map (fun _ => approximate (sin_uc 1) (eps 10)) (N.enum 10)).
Time Eval vm_compute in approximate (approx' (AQ := bigD) ARsin_uc 1 0 1 (eps 8)) (eps 8).
Time Eval vm_compute in approximate (simpson_integral sin_uc 1 0 1) (1#100000000)%Qpos.
Time Eval vm_compute in answer 8 (simpson_integral sin_uc 1 0 1).
(*Eval vm_compute in approximate (simpson' (AQ := bigD) ARsin_uc 1 1 (1#1)%Qpos 0) (1#1)%Qpos.*)
(*Eval vm_compute in (*cast _ Q*)
(approximate (approx' (AQ := bigD) ARsin_uc 1 0 1 (1#10)%Qpos) (1#10)%Qpos).*)
Time Eval vm_compute in
cast _ Q (approximate (simpson_integral' (AQ := bigD) ARsin_uc 1 0 1) (1#100000000)%Qpos).
Time Eval vm_compute in N.enum ((2 : nat)^(N' (AQ := bigD) 1 1 (1#10000000000)%Qpos)).
|