| author | huffman | 
| Thu, 07 Jun 2007 02:34:37 +0200 | |
| changeset 23286 | 85e7e043b980 | 
| parent 22384 | 33a46e6c7f04 | 
| child 24742 | 73b8b42a36b6 | 
| permissions | -rw-r--r-- | 
| 7998 | 1 | (* | 
| 2 | Experimental theory: long division of polynomials | |
| 3 | $Id$ | |
| 4 | Author: Clemens Ballarin, started 23 June 1999 | |
| 5 | *) | |
| 6 | ||
| 16417 | 7 | theory LongDiv imports PolyHomo begin | 
| 7998 | 8 | |
| 21423 | 9 | definition | 
| 10 | lcoeff :: "'a::ring up => 'a" where | |
| 11 | "lcoeff p = coeff p (deg p)" | |
| 7998 | 12 | |
| 21423 | 13 | definition | 
| 14 | eucl_size :: "'a::zero up => nat" where | |
| 15 | "eucl_size p = (if p = 0 then 0 else deg p + 1)" | |
| 14723 | 16 | |
| 17 | lemma SUM_shrink_below_lemma: | |
| 18 | "!! f::(nat=>'a::ring). (ALL i. i < m --> f i = 0) --> | |
| 19 |   setsum (%i. f (i+m)) {..d} = setsum f {..m+d}"
 | |
| 20 | apply (induct_tac d) | |
| 15481 | 21 | apply (induct_tac m) | 
| 21423 | 22 | apply simp | 
| 23 | apply force | |
| 22384 
33a46e6c7f04
prefix of class interpretation not mandatory any longer
 haftmann parents: 
21423diff
changeset | 24 | apply (simp add: add_commute [of m]) | 
| 21423 | 25 | done | 
| 26 | ||
| 27 | lemma SUM_extend_below: | |
| 28 | "!! f::(nat=>'a::ring). | |
| 29 |      [| m <= n; !!i. i < m ==> f i = 0; P (setsum (%i. f (i+m)) {..n-m}) |]  
 | |
| 30 |      ==> P (setsum f {..n})"
 | |
| 31 | by (simp add: SUM_shrink_below_lemma add_diff_inverse leD) | |
| 32 | ||
| 33 | lemma up_repr2D: | |
| 34 | "!! p::'a::ring up. | |
| 35 |    [| deg p <= n; P (setsum (%i. monom (coeff p i) i) {..n}) |]  
 | |
| 36 | ==> P p" | |
| 37 | by (simp add: up_repr_le) | |
| 38 | ||
| 39 | ||
| 40 | (* Start of LongDiv *) | |
| 41 | ||
| 42 | lemma deg_lcoeff_cancel: | |
| 43 |   "!!p::('a::ring up).  
 | |
| 44 | [| deg p <= deg r; deg q <= deg r; | |
| 45 | coeff p (deg r) = - (coeff q (deg r)); deg r ~= 0 |] ==> | |
| 46 | deg (p + q) < deg r" | |
| 47 | apply (rule_tac j = "deg r - 1" in le_less_trans) | |
| 48 | prefer 2 | |
| 49 | apply arith | |
| 50 | apply (rule deg_aboveI) | |
| 51 | apply (case_tac "deg r = m") | |
| 52 | apply clarify | |
| 53 | apply simp | |
| 54 | (* case "deg q ~= m" *) | |
| 55 | apply (subgoal_tac "deg p < m & deg q < m") | |
| 56 | apply (simp (no_asm_simp) add: deg_aboveD) | |
| 57 | apply arith | |
| 58 | done | |
| 59 | ||
| 60 | lemma deg_lcoeff_cancel2: | |
| 61 |   "!!p::('a::ring up).  
 | |
| 62 | [| deg p <= deg r; deg q <= deg r; | |
| 63 | p ~= -q; coeff p (deg r) = - (coeff q (deg r)) |] ==> | |
| 64 | deg (p + q) < deg r" | |
| 65 | apply (rule deg_lcoeff_cancel) | |
| 66 | apply assumption+ | |
| 67 | apply (rule classical) | |
| 68 | apply clarify | |
| 69 | apply (erule notE) | |
| 70 | apply (rule_tac p = p in up_repr2D, assumption) | |
| 71 | apply (rule_tac p = q in up_repr2D, assumption) | |
| 72 | apply (rotate_tac -1) | |
| 73 | apply (simp add: smult_l_minus) | |
| 74 | done | |
| 75 | ||
| 76 | lemma long_div_eucl_size: | |
| 77 |   "!!g::('a::ring up). g ~= 0 ==>  
 | |
| 78 | Ex (% (q, r, k). | |
| 79 | (lcoeff g)^k *s f = q * g + r & (eucl_size r < eucl_size g))" | |
| 80 | apply (rule_tac P = "%f. Ex (% (q, r, k) . (lcoeff g) ^k *s f = q * g + r & (eucl_size r < eucl_size g))" in wf_induct) | |
| 81 | (* TO DO: replace by measure_induct *) | |
| 82 | apply (rule_tac f = eucl_size in wf_measure) | |
| 83 | apply (case_tac "eucl_size x < eucl_size g") | |
| 84 | apply (rule_tac x = "(0, x, 0)" in exI) | |
| 85 | apply (simp (no_asm_simp)) | |
| 86 | (* case "eucl_size x >= eucl_size g" *) | |
| 87 | apply (drule_tac x = "lcoeff g *s x - (monom (lcoeff x) (deg x - deg g)) * g" in spec) | |
| 88 | apply (erule impE) | |
| 89 | apply (simp (no_asm_use) add: inv_image_def measure_def lcoeff_def) | |
| 90 | apply (case_tac "x = 0") | |
| 91 | apply (rotate_tac -1) | |
| 92 | apply (simp add: eucl_size_def) | |
| 93 | (* case "x ~= 0 *) | |
| 94 | apply (rotate_tac -1) | |
| 95 | apply (simp add: eucl_size_def) | |
| 96 | apply (rule impI) | |
| 97 | apply (rule deg_lcoeff_cancel2) | |
| 98 | (* replace by linear arithmetic??? *) | |
| 99 | apply (rule_tac [2] le_trans) | |
| 100 | apply (rule_tac [2] deg_smult_ring) | |
| 101 | prefer 2 | |
| 102 | apply simp | |
| 103 | apply (simp (no_asm)) | |
| 104 | apply (rule le_trans) | |
| 105 | apply (rule deg_mult_ring) | |
| 106 | apply (rule le_trans) | |
| 107 | (**) | |
| 108 | apply (rule add_le_mono) | |
| 109 | apply (rule le_refl) | |
| 110 | (* term order forces to use this instead of add_le_mono1 *) | |
| 111 | apply (rule deg_monom_ring) | |
| 112 | apply (simp (no_asm_simp)) | |
| 113 | apply force | |
| 114 | apply (simp (no_asm)) | |
| 115 | (**) | |
| 116 | (* This change is probably caused by application of commutativity *) | |
| 117 | apply (rule_tac m = "deg g" and n = "deg x" in SUM_extend) | |
| 118 | apply (simp (no_asm)) | |
| 119 | apply (simp (no_asm_simp)) | |
| 120 | apply arith | |
| 121 | apply (rule_tac m = "deg g" and n = "deg g" in SUM_extend_below) | |
| 122 | apply (rule le_refl) | |
| 123 | apply (simp (no_asm_simp)) | |
| 124 | apply arith | |
| 125 | apply (simp (no_asm)) | |
| 126 | (**) | |
| 127 | (* end of subproof deg f1 < deg f *) | |
| 128 | apply (erule exE) | |
| 129 | apply (rule_tac x = "((% (q,r,k) . (monom (lcoeff g ^ k * lcoeff x) (deg x - deg g) + q)) xa, (% (q,r,k) . r) xa, (% (q,r,k) . Suc k) xa) " in exI) | |
| 130 | apply clarify | |
| 131 | apply (drule sym) | |
| 132 |   apply (tactic {* simp_tac (simpset() addsimps [thm "l_distr", thm "a_assoc"]
 | |
| 133 | delsimprocs [ring_simproc]) 1 *}) | |
| 134 |   apply (tactic {* asm_simp_tac (simpset() delsimprocs [ring_simproc]) 1 *})
 | |
| 135 |   apply (tactic {* simp_tac (simpset () addsimps [thm "minus_def", thm "smult_r_distr",
 | |
| 136 | thm "smult_r_minus", thm "monom_mult_smult", thm "smult_assoc1", thm "smult_assoc2"] | |
| 137 | delsimprocs [ring_simproc]) 1 *}) | |
| 138 | apply simp | |
| 139 | done | |
| 140 | ||
| 141 | ML {* simplify (simpset() addsimps [thm "eucl_size_def"]
 | |
| 142 | delsimprocs [ring_simproc]) (thm "long_div_eucl_size") *} | |
| 143 | ||
| 144 | thm long_div_eucl_size [simplified] | |
| 145 | ||
| 146 | lemma long_div_ring: | |
| 147 |   "!!g::('a::ring up). g ~= 0 ==>  
 | |
| 148 | Ex (% (q, r, k). | |
| 149 | (lcoeff g)^k *s f = q * g + r & (r = 0 | deg r < deg g))" | |
| 150 |   apply (tactic {* forw_inst_tac [("f", "f")]
 | |
| 151 | (simplify (simpset() addsimps [thm "eucl_size_def"] | |
| 152 | delsimprocs [ring_simproc]) (thm "long_div_eucl_size")) 1 *}) | |
| 153 |   apply (tactic {* auto_tac (claset(), simpset() delsimprocs [ring_simproc]) *})
 | |
| 154 | apply (case_tac "aa = 0") | |
| 155 | apply blast | |
| 156 | (* case "aa ~= 0 *) | |
| 157 | apply (rotate_tac -1) | |
| 158 | apply auto | |
| 159 | done | |
| 160 | ||
| 161 | (* Next one fails *) | |
| 162 | lemma long_div_unit: | |
| 163 |   "!!g::('a::ring up). [| g ~= 0; (lcoeff g) dvd 1 |] ==>  
 | |
| 164 | Ex (% (q, r). f = q * g + r & (r = 0 | deg r < deg g))" | |
| 165 | apply (frule_tac f = "f" in long_div_ring) | |
| 166 | apply (erule exE) | |
| 167 | apply (rule_tac x = "((% (q,r,k) . (inverse (lcoeff g ^k) *s q)) x, (% (q,r,k) . inverse (lcoeff g ^k) *s r) x) " in exI) | |
| 168 | apply clarify | |
| 169 | apply (rule conjI) | |
| 170 | apply (drule sym) | |
| 171 |    apply (tactic {* asm_simp_tac
 | |
| 172 | (simpset() addsimps [thm "smult_r_distr" RS sym, thm "smult_assoc2"] | |
| 173 | delsimprocs [ring_simproc]) 1 *}) | |
| 174 | apply (simp (no_asm_simp) add: l_inverse_ring unit_power smult_assoc1 [symmetric]) | |
| 175 | (* degree property *) | |
| 176 | apply (erule disjE) | |
| 177 | apply (simp (no_asm_simp)) | |
| 178 | apply (rule disjI2) | |
| 179 | apply (rule le_less_trans) | |
| 180 | apply (rule deg_smult_ring) | |
| 181 | apply (simp (no_asm_simp)) | |
| 182 | done | |
| 183 | ||
| 184 | lemma long_div_theorem: | |
| 185 |   "!!g::('a::field up). g ~= 0 ==>  
 | |
| 186 | Ex (% (q, r). f = q * g + r & (r = 0 | deg r < deg g))" | |
| 187 | apply (rule long_div_unit) | |
| 188 | apply assumption | |
| 189 | apply (simp (no_asm_simp) add: lcoeff_def lcoeff_nonzero field_ax) | |
| 190 | done | |
| 191 | ||
| 192 | lemma uminus_zero: "- (0::'a::ring) = 0" | |
| 193 | by simp | |
| 194 | ||
| 195 | lemma diff_zero_imp_eq: "!!a::'a::ring. a - b = 0 ==> a = b" | |
| 196 | apply (rule_tac s = "a - (a - b) " in trans) | |
| 197 |    apply (tactic {* asm_simp_tac (simpset() delsimprocs [ring_simproc]) 1 *})
 | |
| 198 | apply simp | |
| 199 | apply (simp (no_asm)) | |
| 200 | done | |
| 201 | ||
| 202 | lemma eq_imp_diff_zero: "!!a::'a::ring. a = b ==> a + (-b) = 0" | |
| 203 | by simp | |
| 204 | ||
| 205 | lemma long_div_quo_unique: | |
| 206 |   "!!g::('a::field up). [| g ~= 0;  
 | |
| 207 | f = q1 * g + r1; (r1 = 0 | deg r1 < deg g); | |
| 208 | f = q2 * g + r2; (r2 = 0 | deg r2 < deg g) |] ==> q1 = q2" | |
| 209 | apply (subgoal_tac "(q1 - q2) * g = r2 - r1") (* 1 *) | |
| 210 | apply (erule_tac V = "f = ?x" in thin_rl) | |
| 211 | apply (erule_tac V = "f = ?x" in thin_rl) | |
| 212 | apply (rule diff_zero_imp_eq) | |
| 213 | apply (rule classical) | |
| 214 | apply (erule disjE) | |
| 215 | (* r1 = 0 *) | |
| 216 | apply (erule disjE) | |
| 217 | (* r2 = 0 *) | |
| 218 |      apply (tactic {* asm_full_simp_tac (simpset()
 | |
| 219 | addsimps [thm "integral_iff", thm "minus_def", thm "l_zero", thm "uminus_zero"] | |
| 220 | delsimprocs [ring_simproc]) 1 *}) | |
| 221 | (* r2 ~= 0 *) | |
| 222 | apply (drule_tac f = "deg" and y = "r2 - r1" in arg_cong) | |
| 223 |     apply (tactic {* asm_full_simp_tac (simpset() addsimps
 | |
| 224 | [thm "minus_def", thm "l_zero", thm "uminus_zero"] delsimprocs [ring_simproc]) 1 *}) | |
| 225 | (* r1 ~=0 *) | |
| 226 | apply (erule disjE) | |
| 227 | (* r2 = 0 *) | |
| 228 | apply (drule_tac f = "deg" and y = "r2 - r1" in arg_cong) | |
| 229 |     apply (tactic {* asm_full_simp_tac (simpset() addsimps
 | |
| 230 | [thm "minus_def", thm "l_zero", thm "uminus_zero"] delsimprocs [ring_simproc]) 1 *}) | |
| 231 | (* r2 ~= 0 *) | |
| 232 | apply (drule_tac f = "deg" and y = "r2 - r1" in arg_cong) | |
| 233 |    apply (tactic {* asm_full_simp_tac (simpset() addsimps [thm "minus_def"]
 | |
| 234 | delsimprocs [ring_simproc]) 1 *}) | |
| 235 | apply (drule order_eq_refl [THEN add_leD2]) | |
| 236 | apply (drule leD) | |
| 237 | apply (erule notE, rule deg_add [THEN le_less_trans]) | |
| 238 | apply (simp (no_asm_simp)) | |
| 239 | (* proof of 1 *) | |
| 240 | apply (rule diff_zero_imp_eq) | |
| 241 | apply hypsubst | |
| 242 | apply (drule_tac a = "?x+?y" in eq_imp_diff_zero) | |
| 243 | apply simp | |
| 244 | done | |
| 245 | ||
| 246 | lemma long_div_rem_unique: | |
| 247 |   "!!g::('a::field up). [| g ~= 0;  
 | |
| 248 | f = q1 * g + r1; (r1 = 0 | deg r1 < deg g); | |
| 249 | f = q2 * g + r2; (r2 = 0 | deg r2 < deg g) |] ==> r1 = r2" | |
| 250 | apply (subgoal_tac "q1 = q2") | |
| 251 | apply clarify | |
| 252 | apply (rule_tac a = "q2 * g + r1 - q2 * g" and b = "q2 * g + r2 - q2 * g" in box_equals) | |
| 253 | apply simp | |
| 254 | apply (simp (no_asm)) | |
| 255 | apply (simp (no_asm)) | |
| 256 | apply (rule long_div_quo_unique) | |
| 257 | apply assumption+ | |
| 14723 | 258 | done | 
| 7998 | 259 | |
| 260 | end |