ErrorBounds.v 16.9 KB
Newer Older
1
(**
Heiko Becker's avatar
Heiko Becker committed
2
3
Proofs of general bounds on the error of arithmetic expressions.
This shortens soundness proofs later.
4
Bounds are explained in section 5, Deriving Computable Error Bounds
5
**)
6
Require Import Coq.Reals.Reals Coq.micromega.Psatz Coq.QArith.QArith Coq.QArith.Qreals.
Heiko Becker's avatar
Heiko Becker committed
7
8
Require Import Daisy.Infra.Abbrevs Daisy.Infra.RationalSimps Daisy.Infra.RealSimps Daisy.Infra.RealRationalProps.
Require Import Daisy.Environments Daisy.Infra.ExpressionAbbrevs.
Heiko Becker's avatar
Heiko Becker committed
9

10
11
12
Lemma const_abs_err_bounded (n:R) (nR:R) (nF:R) (E1 E2:env) (absenv:analysisResult):
  eval_exp 0%R E1 (Const n) nR ->
  eval_exp (Q2R machineEpsilon) E2 (Const n) nF ->
13
  (Rabs (nR - nF) <= Rabs n * (Q2R machineEpsilon))%R.
14
Proof.
Heiko Becker's avatar
Heiko Becker committed
15
  intros eval_real eval_float.
16
  inversion eval_real; subst.
17
  rewrite delta_0_deterministic; auto.
18
19
  inversion eval_float; subst.
  unfold perturb; simpl.
20
  rewrite Rabs_err_simpl, Rabs_mult.
21
22
23
  apply Rmult_le_compat_l; [apply Rabs_pos | auto].
Qed.

24
Lemma add_abs_err_bounded (e1:exp Q) (e1R:R) (e1F:R) (e2:exp Q) (e2R:R) (e2F:R)
25
26
27
28
29
30
31
      (vR:R) (vF:R) (E1 E2:env) (err1 err2 :Q):
  eval_exp 0%R E1 (toRExp e1) e1R ->
  eval_exp (Q2R machineEpsilon) E2 (toRExp e1) e1F ->
  eval_exp 0%R E1 (toRExp e2) e2R ->
  eval_exp (Q2R machineEpsilon) E2 (toRExp e2) e2F ->
  eval_exp 0%R E1 (Binop Plus (toRExp e1) (toRExp e2)) vR ->
  eval_exp (Q2R machineEpsilon) (updEnv 2 e2F (updEnv 1 e1F emptyEnv)) (Binop Plus (Var R 1) (Var R 2)) vF ->
Heiko Becker's avatar
Heiko Becker committed
32
33
34
  (Rabs (e1R - e1F) <= Q2R err1)%R ->
  (Rabs (e2R - e2F) <= Q2R err2)%R ->
  (Rabs (vR - vF) <= Q2R err1 + Q2R err2 + (Rabs (e1F + e2F) * (Q2R machineEpsilon)))%R.
35
Proof.
Heiko Becker's avatar
Heiko Becker committed
36
  intros e1_real e1_float e2_real e2_float plus_real plus_float bound_e1 bound_e2.
37
38
  (* Prove that e1R and e2R are the correct values and that vR is e1R + e2R *)
  inversion plus_real; subst.
39
  rewrite delta_0_deterministic in plus_real; auto.
40
41
  rewrite (delta_0_deterministic (evalBinop Plus v1 v2) delta); auto.
  unfold evalBinop in *; simpl in *.
42
  clear delta H2.
43
  rewrite (meps_0_deterministic H3 e1_real);
44
    rewrite (meps_0_deterministic H5 e2_real).
45
  rewrite (meps_0_deterministic H3 e1_real) in plus_real.
46
  rewrite (meps_0_deterministic H5 e2_real) in plus_real.
47
  clear H3 H5 H6 v1 v2.
48
49
50
  (* Now unfold the float valued evaluation to get the deltas we need for the inequality *)
  inversion plus_float; subst.
  unfold perturb; simpl.
51
  inversion H3; subst; inversion H5; subst.
52
  unfold updEnv; simpl.
53
54
55
  unfold updEnv in H0,H1; simpl in *.
  symmetry in H0, H1.
  inversion H0; inversion H1; subst.
56
  (* We have now obtained all necessary values from the evaluations --> remove them for readability *)
57
  clear plus_float H3 H5 plus_real e1_real e1_float e2_real e2_float H0 H1.
58
59
60
61
62
63
64
65
66
67
68
69
  repeat rewrite Rmult_plus_distr_l.
  rewrite Rmult_1_r.
  rewrite Rsub_eq_Ropp_Rplus.
  repeat rewrite Ropp_plus_distr.
  rewrite plus_bounds_simplify.
  pose proof (Rabs_triang (e1R + - e1F) ((e2R + - e2F) + - ((e1F + e2F) * delta))).
  rewrite Rplus_assoc.
  eapply Rle_trans.
  apply H.
  pose proof (Rabs_triang (e2R + - e2F) (- ((e1F + e2F) * delta))).
  pose proof (Rplus_le_compat_l (Rabs (e1R + - e1F)) _ _ H0).
  eapply Rle_trans.
70
  apply H1.
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
  rewrite <- Rplus_assoc.
  repeat rewrite <- Rsub_eq_Ropp_Rplus.
  rewrite Rabs_Ropp.
  eapply Rplus_le_compat.
  - eapply Rplus_le_compat; auto.
  - rewrite Rabs_mult.
    eapply Rle_trans.
    eapply Rmult_le_compat_l.
    apply Rabs_pos.
    apply H2.
    apply Req_le; auto.
Qed.

(**
  Copy-Paste proof with minor differences, was easier then manipulating the evaluations and then applying the lemma
**)
87
Lemma subtract_abs_err_bounded (e1:exp Q) (e1R:R) (e1F:R) (e2:exp Q) (e2R:R)
88
89
90
91
92
93
94
      (e2F:R) (vR:R) (vF:R) (E1 E2:env) err1 err2:
  eval_exp 0%R E1 (toRExp e1) e1R ->
  eval_exp (Q2R machineEpsilon) E2   (toRExp e1) e1F ->
  eval_exp 0%R E1 (toRExp e2) e2R ->
  eval_exp (Q2R machineEpsilon) E2 (toRExp e2) e2F ->
  eval_exp 0%R E1 (Binop Sub (toRExp e1) (toRExp e2)) vR ->
  eval_exp (Q2R machineEpsilon) (updEnv 2 e2F (updEnv 1 e1F emptyEnv)) (Binop Sub (Var R 1) (Var R 2)) vF ->
Heiko Becker's avatar
Heiko Becker committed
95
96
97
  (Rabs (e1R - e1F) <= Q2R err1)%R ->
  (Rabs (e2R - e2F) <= Q2R err2)%R ->
  (Rabs (vR - vF) <= Q2R err1 + Q2R err2 + ((Rabs (e1F - e2F)) * (Q2R machineEpsilon)))%R.
98
Proof.
Heiko Becker's avatar
Heiko Becker committed
99
  intros e1_real e1_float e2_real e2_float sub_real sub_float bound_e1 bound_e2.
100
101
  (* Prove that e1R and e2R are the correct values and that vR is e1R + e2R *)
  inversion sub_real; subst.
102
103
  rewrite delta_0_deterministic in sub_real; auto.
  rewrite delta_0_deterministic; auto.
104
  unfold evalBinop in *; simpl in *.
105
  clear delta H2.
106
  rewrite (meps_0_deterministic H3 e1_real);
107
    rewrite (meps_0_deterministic H5 e2_real).
108
  rewrite (meps_0_deterministic H3 e1_real) in sub_real.
109
  rewrite (meps_0_deterministic H5 e2_real) in sub_real.
110
  clear H3 H5 H6 v1 v2.
111
112
113
  (* Now unfold the float valued evaluation to get the deltas we need for the inequality *)
  inversion sub_float; subst.
  unfold perturb; simpl.
114
  inversion H3; subst; inversion H5; subst.
115
  unfold updEnv; simpl.
116
117
118
  symmetry in H0, H1.
  unfold updEnv in H0, H1; simpl in H0, H1.
  inversion H0; inversion H1; subst.
119
  (* We have now obtained all necessary values from the evaluations --> remove them for readability *)
120
  clear sub_float H3 H5 sub_real e1_real e1_float e2_real e2_float H0 H1.
121
122
123
124
125
  repeat rewrite Rmult_plus_distr_l.
  rewrite Rmult_1_r.
  repeat rewrite Rsub_eq_Ropp_Rplus.
  repeat rewrite Ropp_plus_distr.
  rewrite plus_bounds_simplify.
126
  rewrite Ropp_involutive.
127
128
  rewrite Rplus_assoc.
  eapply Rle_trans.
129
  apply Rabs_triang.
130
  eapply Rle_trans.
131
132
  eapply Rplus_le_compat_l.
  apply Rabs_triang.
133
  rewrite <- Rplus_assoc.
134
  setoid_rewrite Rplus_comm at 4.
135
136
  repeat rewrite <- Rsub_eq_Ropp_Rplus.
  rewrite Rabs_Ropp.
137
138
139
140
  rewrite Rabs_minus_sym in bound_e2.
  apply Rplus_le_compat; [apply Rplus_le_compat; auto | ].
  rewrite Rabs_mult.
  eapply Rmult_le_compat_l; [apply Rabs_pos | auto].
141
142
Qed.

143
Lemma mult_abs_err_bounded (e1:exp Q) (e1R:R) (e1F:R) (e2:exp Q) (e2R:R) (e2F:R)
144
145
146
147
148
149
150
      (vR:R) (vF:R) (E1 E2:env):
  eval_exp 0%R E1 (toRExp e1) e1R ->
  eval_exp (Q2R machineEpsilon) E2 (toRExp e1) e1F ->
  eval_exp 0%R E1 (toRExp e2) e2R ->
  eval_exp (Q2R machineEpsilon) E2  (toRExp e2) e2F ->
  eval_exp 0%R E1  (Binop Mult (toRExp e1) (toRExp e2)) vR ->
  eval_exp (Q2R machineEpsilon) (updEnv 2 e2F (updEnv 1 e1F emptyEnv)) (Binop Mult (Var R 1) (Var R 2)) vF ->
151
  (Rabs (vR - vF) <= Rabs (e1R * e2R - e1F * e2F) + Rabs (e1F * e2F) * (Q2R machineEpsilon))%R.
152
Proof.
Heiko Becker's avatar
Heiko Becker committed
153
  intros e1_real e1_float e2_real e2_float mult_real mult_float.
154
  (* Prove that e1R and e2R are the correct values and that vR is e1R * e2R *)
155
  inversion mult_real; subst.
156
157
  rewrite delta_0_deterministic in mult_real; auto.
  rewrite delta_0_deterministic; auto.
158
  unfold evalBinop in *; simpl in *.
159
  clear delta H2.
160
  rewrite (meps_0_deterministic H3 e1_real);
161
    rewrite (meps_0_deterministic H5 e2_real).
162
  rewrite (meps_0_deterministic H3 e1_real) in mult_real.
163
  rewrite (meps_0_deterministic H5 e2_real) in mult_real.
164
  clear H3 H5 H6 v1 v2.
165
166
167
  (* Now unfold the float valued evaluation to get the deltas we need for the inequality *)
  inversion mult_float; subst.
  unfold perturb; simpl.
168
  inversion H3; subst; inversion H5; subst.
169
  symmetry in H0, H1;
170
    unfold updEnv in *; simpl in *.
171
  inversion H0; inversion H1; subst.
172
  (* We have now obtained all necessary values from the evaluations --> remove them for readability *)
173
  clear mult_float H3 H5 mult_real e1_real e1_float e2_real e2_float H0 H1.
174
175
176
177
  repeat rewrite Rmult_plus_distr_l.
  rewrite Rmult_1_r.
  rewrite Rsub_eq_Ropp_Rplus.
  rewrite Ropp_plus_distr.
Heiko Becker's avatar
Heiko Becker committed
178
179
  rewrite <- Rplus_assoc.
  setoid_rewrite <- Rsub_eq_Ropp_Rplus at 2.
180
181
182
  eapply Rle_trans.
  eapply Rabs_triang.
  eapply Rplus_le_compat_l.
Heiko Becker's avatar
Heiko Becker committed
183
  rewrite Rabs_Ropp.
184
  repeat rewrite Rabs_mult.
Heiko Becker's avatar
Heiko Becker committed
185
  eapply Rmult_le_compat_l; auto.
186
187
  rewrite <- Rabs_mult.
  apply Rabs_pos.
188
189
Qed.

190
Lemma div_abs_err_bounded (e1:exp Q) (e1R:R) (e1F:R) (e2:exp Q) (e2R:R) (e2F:R)
191
192
193
194
195
196
197
      (vR:R) (vF:R) (E1 E2:env):
  eval_exp 0%R E1 (toRExp e1) e1R ->
  eval_exp (Q2R machineEpsilon) E2 (toRExp e1) e1F ->
  eval_exp 0%R E1 (toRExp e2) e2R ->
  eval_exp (Q2R machineEpsilon) E2  (toRExp e2) e2F ->
  eval_exp 0%R E1 (Binop Div (toRExp e1) (toRExp e2)) vR ->
  eval_exp (Q2R machineEpsilon) (updEnv 2 e2F (updEnv 1 e1F emptyEnv)) (Binop Div (Var R 1) (Var R 2)) vF ->
198
  (Rabs (vR - vF) <= Rabs (e1R / e2R - e1F / e2F) + Rabs (e1F / e2F) * (Q2R machineEpsilon))%R.
199
Proof.
Heiko Becker's avatar
Heiko Becker committed
200
  intros e1_real e1_float e2_real e2_float div_real div_float.
201
202
  (* Prove that e1R and e2R are the correct values and that vR is e1R * e2R *)
  inversion div_real; subst.
203
204
  rewrite delta_0_deterministic in div_real; auto.
  rewrite delta_0_deterministic; auto.
205
  unfold evalBinop in *; simpl in *.
206
  clear delta H2.
207
  rewrite (meps_0_deterministic H3 e1_real);
208
    rewrite (meps_0_deterministic H5 e2_real).
209
  rewrite (meps_0_deterministic H3 e1_real) in div_real.
210
  rewrite (meps_0_deterministic H5 e2_real) in div_real.
211
  clear H3 H5 H6 v1 v2.
212
213
214
  (* Now unfold the float valued evaluation to get the deltas we need for the inequality *)
  inversion div_float; subst.
  unfold perturb; simpl.
215
  inversion H3; subst; inversion H5; subst.
216
  symmetry in H0, H1;
217
    unfold updEnv in *; simpl in *.
218
  inversion H0; inversion H1; subst.
219
  (* We have now obtained all necessary values from the evaluations --> remove them for readability *)
220
  clear div_float H3 H5 div_real e1_real e1_float e2_real e2_float H0 H1.
221
222
223
224
225
226
227
228
229
230
231
232
233
  repeat rewrite Rmult_plus_distr_l.
  rewrite Rmult_1_r.
  rewrite Rsub_eq_Ropp_Rplus.
  rewrite Ropp_plus_distr.
  rewrite <- Rplus_assoc.
  setoid_rewrite <- Rsub_eq_Ropp_Rplus at 2.
  eapply Rle_trans.
  eapply Rabs_triang.
  eapply Rplus_le_compat_l.
  rewrite Rabs_Ropp.
  repeat rewrite Rabs_mult.
  eapply Rmult_le_compat_l; auto.
  apply Rabs_pos.
234
235
Qed.

236
237
238
239
240
241
242
243
Lemma err_prop_inversion_pos_real nF nR err elo ehi
      (float_iv_pos : (0 < elo - err)%R)
      (real_iv_pos : (0 < elo)%R)
      (err_bounded : (Rabs (nR - nF) <= err)%R)
      (valid_bounds_e2 : (elo <= nR <= ehi)%R)
      (valid_bounds_e2_err : (elo - err <= nF <= ehi + err)%R)
      (err_pos : (0 <= err)%R):
  (Rabs (/nR - / nF) <= err * / ((elo - err) * (elo- err)))%R.
244
Proof.
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
  rewrite Rabs_case_inverted in err_bounded.
  assert (0 < nF)%R as nF_pos by lra.
  destruct err_bounded as [ [diff_pos err_bounded] | [diff_neg err_bounded]].
  - cut (0 < /nF - / nR)%R.
    + intros abs_neg.
      rewrite Rabs_left; try lra.
      rewrite Rsub_eq_Ropp_Rplus, Ropp_plus_distr, Ropp_involutive.
      rewrite Ropp_inv_permute; try lra.
      apply (Rle_trans _ (/ - nR + / (nR - err))).
      * apply Rplus_le_compat_l.
        apply Rinv_le_contravar; lra.
      * rewrite equal_naming_inv; try lra.
        assert (- nR + (nR - err) = - err)%R as simplify_up by lra.
        rewrite simplify_up.
        unfold Rdiv.
        repeat(rewrite <- Ropp_mult_distr_l); rewrite <- Ropp_inv_permute.
        {
          rewrite <- Ropp_mult_distr_r, Ropp_involutive.
          apply Rmult_le_compat_l; try lra.
          apply Rinv_le_contravar.
          - apply Rmult_0_lt_preserving; lra.
          - apply Rmult_le_compat; lra.
        }
        { assert (0 < nR * (nR - err))%R by (apply Rmult_0_lt_preserving; lra); lra. }
    + cut (/ nR < /nF)%R.
      * intros; lra.
      * apply Rinv_lt_contravar; try lra.
        apply Rmult_0_lt_preserving; lra.
  - cut (0 <= /nR - /nF)%R.
    + intros abs_pos.
      rewrite Rabs_right; try lra.
      rewrite Rsub_eq_Ropp_Rplus, Ropp_plus_distr, Ropp_involutive in err_bounded.
      rewrite Rsub_eq_Ropp_Rplus.
      apply (Rle_trans _ (/nR - / (nR + err))).
      * apply Rplus_le_compat_l.
        apply Ropp_le_contravar.
        apply Rinv_le_contravar; lra.
      * rewrite Rsub_eq_Ropp_Rplus, Ropp_inv_permute; try lra.
        rewrite equal_naming_inv; try lra.
        assert (nR + - (nR + err) = - err)%R as simpl_up by lra.
        rewrite simpl_up.
        unfold Rdiv.
        rewrite <- Ropp_mult_distr_l, <- Ropp_mult_distr_r, <- Ropp_inv_permute.
        { rewrite <- Ropp_mult_distr_r. rewrite Ropp_involutive.
          apply Rmult_le_compat_l; try auto.
          apply Rinv_le_contravar.
          - apply Rmult_0_lt_preserving; lra.
          - apply Rmult_le_compat; lra.
        }
        { assert (0 < nR * (nR + err))%R by (apply Rmult_0_lt_preserving; lra); lra. }
    + cut (/nF <= /nR)%R.
      * intros; lra.
      * apply Rinv_le_contravar; try lra.
Qed.

Lemma err_prop_inversion_pos nF nR err (elo ehi:Q)
      (float_iv_pos : (Q2R 0 < Q2R (elo - err))%R)
      (real_iv_pos : (Q2R 0 < Q2R elo)%R)
      (err_bounded : (Rabs (nR - nF) <= Q2R err)%R)
      (valid_bounds_e2 : (Q2R elo <= nR <= Q2R ehi)%R)
      (valid_bounds_e2_err : (Q2R elo - Q2R err <= nF <= Q2R ehi + Q2R err)%R)
      (err_pos : (0 <= Q2R err)%R):
  (Rabs (/nR - / nF) <= Q2R err * / ((Q2R elo- Q2R err) * (Q2R elo- Q2R err)))%R.
Proof.
  eapply err_prop_inversion_pos_real; try rewrite <- Q2R0_is_0; eauto.
  rewrite <- Q2R_minus; auto.
  rewrite Q2R0_is_0; auto.
Qed.

Lemma err_prop_inversion_neg_real nF nR err elo ehi
      (float_iv_neg : (ehi + err < 0)%R)
      (real_iv_neg : (ehi < 0)%R)
      (err_bounded : (Rabs (nR - nF) <= err)%R)
      (valid_bounds_e : (elo <= nR <= ehi)%R)
      (valid_bounds_e_err : (elo - err <= nF <= ehi + err)%R)
      (err_pos : (0 <= err)%R):
  (Rabs (/nR - / nF) <= err * / ((ehi + err) * (ehi + err)))%R.
Proof.
  rewrite Rabs_case_inverted in err_bounded.
  assert (nF < 0)%R as nF_neg by lra.
  destruct err_bounded as [ [diff_pos err_bounded] | [diff_neg err_bounded]].
  - cut (0 < /nF - / nR)%R.
    + intros abs_neg.
      rewrite Rabs_left; try lra.
      rewrite Rsub_eq_Ropp_Rplus, Ropp_plus_distr, Ropp_involutive.
      rewrite Ropp_inv_permute; try lra.
      apply (Rle_trans _ (/ - nR + / (nR - err))).
      * apply Rplus_le_compat_l.
        assert (0 < - nF)%R by lra.
        assert (0 < - (nR - err))%R by lra.
        assert (nR - err <= nF)%R as nR_lower by lra.
        apply Ropp_le_contravar in nR_lower.
        apply Rinv_le_contravar in nR_lower; try lra.
        repeat (rewrite <- Ropp_inv_permute in nR_lower; try lra).
      * rewrite equal_naming_inv; try lra.
        assert (- nR + (nR - err) = - err)%R as simplify_up by lra.
        rewrite simplify_up.
        unfold Rdiv.
        repeat(rewrite <- Ropp_mult_distr_l); rewrite <- Ropp_inv_permute.
        {
          rewrite <- Ropp_mult_distr_r, Ropp_involutive.
          apply Rmult_le_compat_l; try lra.
          apply Rinv_le_contravar.
          - apply Rmult_lt_0_inverting; lra.
          - eapply Rle_trans.
            eapply Rmult_le_compat_neg_l; try lra.
            instantiate (1 := (nR - err)%R); try lra.
            setoid_rewrite Rmult_comm.
            eapply Rmult_le_compat_neg_l; lra.
        }
        { assert (0 < nR * (nR - err))%R by (apply Rmult_lt_0_inverting; lra); lra. }
    + cut (/ nR < /nF)%R.
      * intros; lra.
      * apply Rinv_lt_contravar; try lra.
        apply Rmult_lt_0_inverting; lra.
  - cut (0 <= /nR - /nF)%R.
    + intros abs_pos.
      rewrite Rabs_right; try lra.
      rewrite Rsub_eq_Ropp_Rplus, Ropp_plus_distr, Ropp_involutive in err_bounded.
      rewrite Rsub_eq_Ropp_Rplus.
      apply (Rle_trans _ (/nR - / (nR + err))).
      * apply Rplus_le_compat_l.
        apply Ropp_le_contravar.
        assert (0 < - nF)%R by lra.
        assert (0 < - (nR + err))%R by lra.
        assert (nF <= nR + err)%R as nR_upper by lra.
        apply Ropp_le_contravar in nR_upper.
        apply Rinv_le_contravar in nR_upper; try lra.
        repeat (rewrite <- Ropp_inv_permute in nR_upper; try lra).
      * rewrite Rsub_eq_Ropp_Rplus, Ropp_inv_permute; try lra.
375
        rewrite equal_naming_inv; try lra.
376
377
        assert (nR + - (nR + err) = - err)%R as simpl_up by lra.
        rewrite simpl_up.
378
        unfold Rdiv.
379
        rewrite <- Ropp_mult_distr_l, <- Ropp_mult_distr_r, <- Ropp_inv_permute.
380
381
382
        { rewrite <- Ropp_mult_distr_r. rewrite Ropp_involutive.
          apply Rmult_le_compat_l; try auto.
          apply Rinv_le_contravar.
383
384
385
386
387
388
389
390
391
392
393
394
395
396
          - apply Rmult_lt_0_inverting; lra.
          - eapply Rle_trans.
            eapply Rmult_le_compat_neg_l; try lra.
            instantiate (1:= (nR + err)%R); try lra.
            setoid_rewrite Rmult_comm.
            eapply Rmult_le_compat_neg_l; lra.
        }
        { assert (0 < nR * (nR + err))%R by (apply Rmult_lt_0_inverting; lra); lra. }
    + cut (/nF <= /nR)%R.
      * intros; lra.
      * assert (nR <= nF)%R by lra.
        assert (- nF <= - nR)%R as le_inv by lra.
        apply Rinv_le_contravar in le_inv; try lra.
        repeat (rewrite <- Ropp_inv_permute in le_inv; try lra).
397
398
Qed.

399
400
401
402
403
404
405
406
Lemma err_prop_inversion_neg nF nR err (elo ehi:Q)
      (float_iv_neg : (Q2R (ehi + err) < Q2R 0)%R)
      (real_iv_neg : (Q2R ehi < Q2R 0)%R)
      (err_bounded : (Rabs (nR - nF) <= Q2R err)%R)
      (valid_bounds_e : (Q2R elo <= nR <= Q2R ehi)%R)
      (valid_bounds_e_err : (Q2R elo - Q2R err <= nF <= Q2R ehi + Q2R err)%R)
      (err_pos : (0 <= Q2R err)%R):
  (Rabs (/nR - / nF) <= Q2R err * / ((Q2R ehi + Q2R err) * (Q2R ehi + Q2R err)))%R.
407
Proof.
408
409
410
411
412
413
  eapply err_prop_inversion_neg_real; try rewrite <- Q2R0_is_0; try lra.
  rewrite <- Q2R_plus ; auto.
  apply valid_bounds_e.
  auto.
  rewrite Q2R0_is_0; auto.
Qed.