(visit-tags-table-buffer): Add autoload cookie;
[emacs.git] / lisp / calc / calc-math.el
blob093679232e591b3f2b839e4d3c2869d5b469ea10
1 ;;; calc-math.el --- mathematical functions for Calc
3 ;; Copyright (C) 1990, 1991, 1992, 1993, 2001 Free Software Foundation, Inc.
5 ;; Author: David Gillespie <daveg@synaptics.com>
6 ;; Maintainers: D. Goel <deego@gnufans.org>
7 ;; Colin Walters <walters@debian.org>
9 ;; This file is part of GNU Emacs.
11 ;; GNU Emacs is distributed in the hope that it will be useful,
12 ;; but WITHOUT ANY WARRANTY. No author or distributor
13 ;; accepts responsibility to anyone for the consequences of using it
14 ;; or for whether it serves any particular purpose or works at all,
15 ;; unless he says so in writing. Refer to the GNU Emacs General Public
16 ;; License for full details.
18 ;; Everyone is granted permission to copy, modify and redistribute
19 ;; GNU Emacs, but only under the conditions described in the
20 ;; GNU Emacs General Public License. A copy of this license is
21 ;; supposed to have been given to you along with GNU Emacs so you
22 ;; can know your rights and responsibilities. It should be in a
23 ;; file named COPYING. Among other things, the copyright notice
24 ;; and this notice must be preserved on all copies.
26 ;;; Commentary:
28 ;;; Code:
30 ;; This file is autoloaded from calc-ext.el.
31 (require 'calc-ext)
33 (require 'calc-macs)
35 (defun calc-Need-calc-math () nil)
38 (defun calc-sqrt (arg)
39 (interactive "P")
40 (calc-slow-wrapper
41 (if (calc-is-inverse)
42 (calc-unary-op "^2" 'calcFunc-sqr arg)
43 (calc-unary-op "sqrt" 'calcFunc-sqrt arg))))
45 (defun calc-isqrt (arg)
46 (interactive "P")
47 (calc-slow-wrapper
48 (if (calc-is-inverse)
49 (calc-unary-op "^2" 'calcFunc-sqr arg)
50 (calc-unary-op "isqt" 'calcFunc-isqrt arg))))
53 (defun calc-hypot (arg)
54 (interactive "P")
55 (calc-slow-wrapper
56 (calc-binary-op "hypt" 'calcFunc-hypot arg)))
58 (defun calc-ln (arg)
59 (interactive "P")
60 (calc-invert-func)
61 (calc-exp arg))
63 (defun calc-log10 (arg)
64 (interactive "P")
65 (calc-hyperbolic-func)
66 (calc-ln arg))
68 (defun calc-log (arg)
69 (interactive "P")
70 (calc-slow-wrapper
71 (if (calc-is-inverse)
72 (calc-binary-op "alog" 'calcFunc-alog arg)
73 (calc-binary-op "log" 'calcFunc-log arg))))
75 (defun calc-ilog (arg)
76 (interactive "P")
77 (calc-slow-wrapper
78 (if (calc-is-inverse)
79 (calc-binary-op "alog" 'calcFunc-alog arg)
80 (calc-binary-op "ilog" 'calcFunc-ilog arg))))
82 (defun calc-lnp1 (arg)
83 (interactive "P")
84 (calc-invert-func)
85 (calc-expm1 arg))
87 (defun calc-exp (arg)
88 (interactive "P")
89 (calc-slow-wrapper
90 (if (calc-is-hyperbolic)
91 (if (calc-is-inverse)
92 (calc-unary-op "lg10" 'calcFunc-log10 arg)
93 (calc-unary-op "10^" 'calcFunc-exp10 arg))
94 (if (calc-is-inverse)
95 (calc-unary-op "ln" 'calcFunc-ln arg)
96 (calc-unary-op "exp" 'calcFunc-exp arg)))))
98 (defun calc-expm1 (arg)
99 (interactive "P")
100 (calc-slow-wrapper
101 (if (calc-is-inverse)
102 (calc-unary-op "ln+1" 'calcFunc-lnp1 arg)
103 (calc-unary-op "ex-1" 'calcFunc-expm1 arg))))
105 (defun calc-pi ()
106 (interactive)
107 (calc-slow-wrapper
108 (if (calc-is-inverse)
109 (if (calc-is-hyperbolic)
110 (if calc-symbolic-mode
111 (calc-pop-push-record 0 "phi" '(var phi var-phi))
112 (calc-pop-push-record 0 "phi" (math-phi)))
113 (if calc-symbolic-mode
114 (calc-pop-push-record 0 "gmma" '(var gamma var-gamma))
115 (calc-pop-push-record 0 "gmma" (math-gamma-const))))
116 (if (calc-is-hyperbolic)
117 (if calc-symbolic-mode
118 (calc-pop-push-record 0 "e" '(var e var-e))
119 (calc-pop-push-record 0 "e" (math-e)))
120 (if calc-symbolic-mode
121 (calc-pop-push-record 0 "pi" '(var pi var-pi))
122 (calc-pop-push-record 0 "pi" (math-pi)))))))
124 (defun calc-sin (arg)
125 (interactive "P")
126 (calc-slow-wrapper
127 (if (calc-is-hyperbolic)
128 (if (calc-is-inverse)
129 (calc-unary-op "asnh" 'calcFunc-arcsinh arg)
130 (calc-unary-op "sinh" 'calcFunc-sinh arg))
131 (if (calc-is-inverse)
132 (calc-unary-op "asin" 'calcFunc-arcsin arg)
133 (calc-unary-op "sin" 'calcFunc-sin arg)))))
135 (defun calc-arcsin (arg)
136 (interactive "P")
137 (calc-invert-func)
138 (calc-sin arg))
140 (defun calc-sinh (arg)
141 (interactive "P")
142 (calc-hyperbolic-func)
143 (calc-sin arg))
145 (defun calc-arcsinh (arg)
146 (interactive "P")
147 (calc-invert-func)
148 (calc-hyperbolic-func)
149 (calc-sin arg))
151 (defun calc-cos (arg)
152 (interactive "P")
153 (calc-slow-wrapper
154 (if (calc-is-hyperbolic)
155 (if (calc-is-inverse)
156 (calc-unary-op "acsh" 'calcFunc-arccosh arg)
157 (calc-unary-op "cosh" 'calcFunc-cosh arg))
158 (if (calc-is-inverse)
159 (calc-unary-op "acos" 'calcFunc-arccos arg)
160 (calc-unary-op "cos" 'calcFunc-cos arg)))))
162 (defun calc-arccos (arg)
163 (interactive "P")
164 (calc-invert-func)
165 (calc-cos arg))
167 (defun calc-cosh (arg)
168 (interactive "P")
169 (calc-hyperbolic-func)
170 (calc-cos arg))
172 (defun calc-arccosh (arg)
173 (interactive "P")
174 (calc-invert-func)
175 (calc-hyperbolic-func)
176 (calc-cos arg))
178 (defun calc-sincos ()
179 (interactive)
180 (calc-slow-wrapper
181 (if (calc-is-inverse)
182 (calc-enter-result 1 "asnc" (list 'calcFunc-arcsincos (calc-top-n 1)))
183 (calc-enter-result 1 "sncs" (list 'calcFunc-sincos (calc-top-n 1))))))
185 (defun calc-tan (arg)
186 (interactive "P")
187 (calc-slow-wrapper
188 (if (calc-is-hyperbolic)
189 (if (calc-is-inverse)
190 (calc-unary-op "atnh" 'calcFunc-arctanh arg)
191 (calc-unary-op "tanh" 'calcFunc-tanh arg))
192 (if (calc-is-inverse)
193 (calc-unary-op "atan" 'calcFunc-arctan arg)
194 (calc-unary-op "tan" 'calcFunc-tan arg)))))
196 (defun calc-arctan (arg)
197 (interactive "P")
198 (calc-invert-func)
199 (calc-tan arg))
201 (defun calc-tanh (arg)
202 (interactive "P")
203 (calc-hyperbolic-func)
204 (calc-tan arg))
206 (defun calc-arctanh (arg)
207 (interactive "P")
208 (calc-invert-func)
209 (calc-hyperbolic-func)
210 (calc-tan arg))
212 (defun calc-arctan2 ()
213 (interactive)
214 (calc-slow-wrapper
215 (calc-enter-result 2 "atn2" (cons 'calcFunc-arctan2 (calc-top-list-n 2)))))
217 (defun calc-conj (arg)
218 (interactive "P")
219 (calc-wrapper
220 (calc-unary-op "conj" 'calcFunc-conj arg)))
222 (defun calc-imaginary ()
223 (interactive)
224 (calc-slow-wrapper
225 (calc-pop-push-record 1 "i*" (math-imaginary (calc-top-n 1)))))
229 (defun calc-to-degrees (arg)
230 (interactive "P")
231 (calc-wrapper
232 (calc-unary-op ">deg" 'calcFunc-deg arg)))
234 (defun calc-to-radians (arg)
235 (interactive "P")
236 (calc-wrapper
237 (calc-unary-op ">rad" 'calcFunc-rad arg)))
240 (defun calc-degrees-mode (arg)
241 (interactive "p")
242 (cond ((= arg 1)
243 (calc-wrapper
244 (calc-change-mode 'calc-angle-mode 'deg)
245 (message "Angles measured in degrees")))
246 ((= arg 2) (calc-radians-mode))
247 ((= arg 3) (calc-hms-mode))
248 (t (error "Prefix argument out of range"))))
250 (defun calc-radians-mode ()
251 (interactive)
252 (calc-wrapper
253 (calc-change-mode 'calc-angle-mode 'rad)
254 (message "Angles measured in radians")))
257 ;;; Compute the integer square-root floor(sqrt(A)). A > 0. [I I] [Public]
258 ;;; This method takes advantage of the fact that Newton's method starting
259 ;;; with an overestimate always works, even using truncating integer division!
260 (defun math-isqrt (a)
261 (cond ((Math-zerop a) a)
262 ((not (math-natnump a))
263 (math-reject-arg a 'natnump))
264 ((integerp a)
265 (math-isqrt-small a))
267 (math-normalize (cons 'bigpos (cdr (math-isqrt-bignum (cdr a))))))))
269 (defun calcFunc-isqrt (a)
270 (if (math-realp a)
271 (math-isqrt (math-floor a))
272 (math-floor (math-sqrt a))))
275 ;;; This returns (flag . result) where the flag is t if A is a perfect square.
276 (defun math-isqrt-bignum (a) ; [P.l L]
277 (let ((len (length a)))
278 (if (= (% len 2) 0)
279 (let* ((top (nthcdr (- len 2) a)))
280 (math-isqrt-bignum-iter
282 (math-scale-bignum-3
283 (math-bignum-big
284 (1+ (math-isqrt-small
285 (+ (* (nth 1 top) 1000) (car top)))))
286 (1- (/ len 2)))))
287 (let* ((top (nth (1- len) a)))
288 (math-isqrt-bignum-iter
290 (math-scale-bignum-3
291 (list (1+ (math-isqrt-small top)))
292 (/ len 2)))))))
294 (defun math-isqrt-bignum-iter (a guess) ; [l L l]
295 (math-working "isqrt" (cons 'bigpos guess))
296 (let* ((q (math-div-bignum a guess))
297 (s (math-add-bignum (car q) guess))
298 (g2 (math-div2-bignum s))
299 (comp (math-compare-bignum g2 guess)))
300 (if (< comp 0)
301 (math-isqrt-bignum-iter a g2)
302 (cons (and (= comp 0)
303 (math-zerop-bignum (cdr q))
304 (= (% (car s) 2) 0))
305 guess))))
307 (defun math-zerop-bignum (a)
308 (and (eq (car a) 0)
309 (progn
310 (while (eq (car (setq a (cdr a))) 0))
311 (null a))))
313 (defun math-scale-bignum-3 (a n) ; [L L S]
314 (while (> n 0)
315 (setq a (cons 0 a)
316 n (1- n)))
319 (defun math-isqrt-small (a) ; A > 0. [S S]
320 (let ((g (cond ((>= a 10000) 1000)
321 ((>= a 100) 100)
322 (t 10)))
324 (while (< (setq g2 (/ (+ g (/ a g)) 2)) g)
325 (setq g g2))
331 ;;; Compute the square root of a number.
332 ;;; [T N] if possible, else [F N] if possible, else [C N]. [Public]
333 (defun math-sqrt (a)
335 (and (Math-zerop a) a)
336 (and (math-known-nonposp a)
337 (math-imaginary (math-sqrt (math-neg a))))
338 (and (integerp a)
339 (let ((sqrt (math-isqrt-small a)))
340 (if (= (* sqrt sqrt) a)
341 sqrt
342 (if calc-symbolic-mode
343 (list 'calcFunc-sqrt a)
344 (math-sqrt-float (math-float a) (math-float sqrt))))))
345 (and (eq (car-safe a) 'bigpos)
346 (let* ((res (math-isqrt-bignum (cdr a)))
347 (sqrt (math-normalize (cons 'bigpos (cdr res)))))
348 (if (car res)
349 sqrt
350 (if calc-symbolic-mode
351 (list 'calcFunc-sqrt a)
352 (math-sqrt-float (math-float a) (math-float sqrt))))))
353 (and (eq (car-safe a) 'frac)
354 (let* ((num-res (math-isqrt-bignum (cdr (Math-bignum-test (nth 1 a)))))
355 (num-sqrt (math-normalize (cons 'bigpos (cdr num-res))))
356 (den-res (math-isqrt-bignum (cdr (Math-bignum-test (nth 2 a)))))
357 (den-sqrt (math-normalize (cons 'bigpos (cdr den-res)))))
358 (if (and (car num-res) (car den-res))
359 (list 'frac num-sqrt den-sqrt)
360 (if calc-symbolic-mode
361 (if (or (car num-res) (car den-res))
362 (math-div (if (car num-res)
363 num-sqrt (list 'calcFunc-sqrt (nth 1 a)))
364 (if (car den-res)
365 den-sqrt (list 'calcFunc-sqrt (nth 2 a))))
366 (list 'calcFunc-sqrt a))
367 (math-sqrt-float (math-float a)
368 (math-div (math-float num-sqrt) den-sqrt))))))
369 (and (eq (car-safe a) 'float)
370 (if calc-symbolic-mode
371 (if (= (% (nth 2 a) 2) 0)
372 (let ((res (math-isqrt-bignum
373 (cdr (Math-bignum-test (nth 1 a))))))
374 (if (car res)
375 (math-make-float (math-normalize
376 (cons 'bigpos (cdr res)))
377 (/ (nth 2 a) 2))
378 (signal 'inexact-result nil)))
379 (signal 'inexact-result nil))
380 (math-sqrt-float a)))
381 (and (eq (car-safe a) 'cplx)
382 (math-with-extra-prec 2
383 (let* ((d (math-abs a))
384 (imag (math-sqrt (math-mul (math-sub d (nth 1 a))
385 '(float 5 -1)))))
386 (list 'cplx
387 (math-sqrt (math-mul (math-add d (nth 1 a)) '(float 5 -1)))
388 (if (math-negp (nth 2 a)) (math-neg imag) imag)))))
389 (and (eq (car-safe a) 'polar)
390 (list 'polar
391 (math-sqrt (nth 1 a))
392 (math-mul (nth 2 a) '(float 5 -1))))
393 (and (eq (car-safe a) 'sdev)
394 (let ((sqrt (math-sqrt (nth 1 a))))
395 (math-make-sdev sqrt
396 (math-div (nth 2 a) (math-mul sqrt 2)))))
397 (and (eq (car-safe a) 'intv)
398 (not (math-negp (nth 2 a)))
399 (math-make-intv (nth 1 a) (math-sqrt (nth 2 a)) (math-sqrt (nth 3 a))))
400 (and (eq (car-safe a) '*)
401 (or (math-known-nonnegp (nth 1 a))
402 (math-known-nonnegp (nth 2 a)))
403 (math-mul (math-sqrt (nth 1 a)) (math-sqrt (nth 2 a))))
404 (and (eq (car-safe a) '/)
405 (or (and (math-known-nonnegp (nth 2 a))
406 (math-div (math-sqrt (nth 1 a)) (math-sqrt (nth 2 a))))
407 (and (math-known-nonnegp (nth 1 a))
408 (not (math-equal-int (nth 1 a) 1))
409 (math-mul (math-sqrt (nth 1 a))
410 (math-sqrt (math-div 1 (nth 2 a)))))))
411 (and (eq (car-safe a) '^)
412 (math-known-evenp (nth 2 a))
413 (math-known-realp (nth 1 a))
414 (math-abs (math-pow (nth 1 a) (math-div (nth 2 a) 2))))
415 (let ((inf (math-infinitep a)))
416 (and inf
417 (math-mul (math-sqrt (math-infinite-dir a inf)) inf)))
418 (progn
419 (calc-record-why 'numberp a)
420 (list 'calcFunc-sqrt a))))
421 (defalias 'calcFunc-sqrt 'math-sqrt)
423 (defun math-infinite-dir (a &optional inf)
424 (or inf (setq inf (math-infinitep a)))
425 (math-normalize (math-expr-subst a inf 1)))
427 (defun math-sqrt-float (a &optional guess) ; [F F F]
428 (if calc-symbolic-mode
429 (signal 'inexact-result nil)
430 (math-with-extra-prec 1 (math-sqrt-raw a guess))))
432 (defun math-sqrt-raw (a &optional guess) ; [F F F]
433 (if (not (Math-posp a))
434 (math-sqrt a)
435 (if (null guess)
436 (let ((ldiff (- (math-numdigs (nth 1 a)) 6)))
437 (or (= (% (+ (nth 2 a) ldiff) 2) 0) (setq ldiff (1+ ldiff)))
438 (setq guess (math-make-float (math-isqrt-small
439 (math-scale-int (nth 1 a) (- ldiff)))
440 (/ (+ (nth 2 a) ldiff) 2)))))
441 (math-sqrt-float-iter a guess)))
443 (defun math-sqrt-float-iter (a guess) ; [F F F]
444 (math-working "sqrt" guess)
445 (let ((g2 (math-mul-float (math-add-float guess (math-div-float a guess))
446 '(float 5 -1))))
447 (if (math-nearly-equal-float g2 guess)
449 (math-sqrt-float-iter a g2))))
451 ;;; True if A and B differ only in the last digit of precision. [P F F]
452 (defun math-nearly-equal-float (a b)
453 (let ((ediff (- (nth 2 a) (nth 2 b))))
454 (cond ((= ediff 0) ;; Expanded out for speed
455 (setq ediff (math-add (Math-integer-neg (nth 1 a)) (nth 1 b)))
456 (or (eq ediff 0)
457 (and (not (consp ediff))
458 (< ediff 10)
459 (> ediff -10)
460 (= (math-numdigs (nth 1 a)) calc-internal-prec))))
461 ((= ediff 1)
462 (setq ediff (math-add (Math-integer-neg (nth 1 b))
463 (math-scale-int (nth 1 a) 1)))
464 (and (not (consp ediff))
465 (< ediff 10)
466 (> ediff -10)
467 (= (math-numdigs (nth 1 b)) calc-internal-prec)))
468 ((= ediff -1)
469 (setq ediff (math-add (Math-integer-neg (nth 1 a))
470 (math-scale-int (nth 1 b) 1)))
471 (and (not (consp ediff))
472 (< ediff 10)
473 (> ediff -10)
474 (= (math-numdigs (nth 1 a)) calc-internal-prec))))))
476 (defun math-nearly-equal (a b) ; [P N N] [Public]
477 (setq a (math-float a))
478 (setq b (math-float b))
479 (if (eq (car a) 'polar) (setq a (math-complex a)))
480 (if (eq (car b) 'polar) (setq b (math-complex b)))
481 (if (eq (car a) 'cplx)
482 (if (eq (car b) 'cplx)
483 (and (or (math-nearly-equal-float (nth 1 a) (nth 1 b))
484 (and (math-nearly-zerop-float (nth 1 a) (nth 2 a))
485 (math-nearly-zerop-float (nth 1 b) (nth 2 b))))
486 (or (math-nearly-equal-float (nth 2 a) (nth 2 b))
487 (and (math-nearly-zerop-float (nth 2 a) (nth 1 a))
488 (math-nearly-zerop-float (nth 2 b) (nth 1 b)))))
489 (and (math-nearly-equal-float (nth 1 a) b)
490 (math-nearly-zerop-float (nth 2 a) b)))
491 (if (eq (car b) 'cplx)
492 (and (math-nearly-equal-float a (nth 1 b))
493 (math-nearly-zerop-float a (nth 2 b)))
494 (math-nearly-equal-float a b))))
496 ;;; True if A is nearly zero compared to B. [P F F]
497 (defun math-nearly-zerop-float (a b)
498 (or (eq (nth 1 a) 0)
499 (<= (+ (math-numdigs (nth 1 a)) (nth 2 a))
500 (1+ (- (+ (math-numdigs (nth 1 b)) (nth 2 b)) calc-internal-prec)))))
502 (defun math-nearly-zerop (a b) ; [P N R] [Public]
503 (setq a (math-float a))
504 (setq b (math-float b))
505 (if (eq (car a) 'cplx)
506 (and (math-nearly-zerop-float (nth 1 a) b)
507 (math-nearly-zerop-float (nth 2 a) b))
508 (if (eq (car a) 'polar)
509 (math-nearly-zerop-float (nth 1 a) b)
510 (math-nearly-zerop-float a b))))
512 ;;; This implementation could be improved, accuracy-wise.
513 (defun math-hypot (a b)
514 (cond ((Math-zerop a) (math-abs b))
515 ((Math-zerop b) (math-abs a))
516 ((not (Math-scalarp a))
517 (if (math-infinitep a)
518 (if (math-infinitep b)
519 (if (equal a b)
521 '(var nan var-nan))
523 (calc-record-why 'scalarp a)
524 (list 'calcFunc-hypot a b)))
525 ((not (Math-scalarp b))
526 (if (math-infinitep b)
528 (calc-record-why 'scalarp b)
529 (list 'calcFunc-hypot a b)))
530 ((and (Math-numberp a) (Math-numberp b))
531 (math-with-extra-prec 1
532 (math-sqrt (math-add (calcFunc-abssqr a) (calcFunc-abssqr b)))))
533 ((eq (car-safe a) 'hms)
534 (if (eq (car-safe b) 'hms) ; this helps sdev's of hms forms
535 (math-to-hms (math-hypot (math-from-hms a 'deg)
536 (math-from-hms b 'deg)))
537 (math-to-hms (math-hypot (math-from-hms a 'deg) b))))
538 ((eq (car-safe b) 'hms)
539 (math-to-hms (math-hypot a (math-from-hms b 'deg))))
540 (t nil)))
541 (defalias 'calcFunc-hypot 'math-hypot)
543 (defun calcFunc-sqr (x)
544 (math-pow x 2))
548 (defun math-nth-root (a n)
549 (cond ((= n 2) (math-sqrt a))
550 ((Math-zerop a) a)
551 ((Math-negp a) nil)
552 ((Math-integerp a)
553 (let ((root (math-nth-root-integer a n)))
554 (if (car root)
555 (cdr root)
556 (and (not calc-symbolic-mode)
557 (math-nth-root-float (math-float a) n
558 (math-float (cdr root)))))))
559 ((eq (car-safe a) 'frac)
560 (let* ((num-root (math-nth-root-integer (nth 1 a) n))
561 (den-root (math-nth-root-integer (nth 2 a) n)))
562 (if (and (car num-root) (car den-root))
563 (list 'frac (cdr num-root) (cdr den-root))
564 (and (not calc-symbolic-mode)
565 (math-nth-root-float
566 (math-float a) n
567 (math-div-float (math-float (cdr num-root))
568 (math-float (cdr den-root))))))))
569 ((eq (car-safe a) 'float)
570 (and (not calc-symbolic-mode)
571 (math-nth-root-float a n)))
572 ((eq (car-safe a) 'polar)
573 (let ((root (math-nth-root (nth 1 a) n)))
574 (and root (list 'polar root (math-div (nth 2 a) n)))))
575 (t nil)))
577 (defun math-nth-root-float (a n &optional guess)
578 (math-inexact-result)
579 (math-with-extra-prec 1
580 (let ((nf (math-float n))
581 (nfm1 (math-float (1- n))))
582 (math-nth-root-float-iter a (or guess
583 (math-make-float
584 1 (/ (+ (math-numdigs (nth 1 a))
585 (nth 2 a)
586 (/ n 2))
587 n)))))))
589 (defun math-nth-root-float-iter (a guess) ; uses "n", "nf", "nfm1"
590 (math-working "root" guess)
591 (let ((g2 (math-div-float (math-add-float (math-mul nfm1 guess)
592 (math-div-float
593 a (math-ipow guess (1- n))))
594 nf)))
595 (if (math-nearly-equal-float g2 guess)
597 (math-nth-root-float-iter a g2))))
599 (defun math-nth-root-integer (a n &optional guess) ; [I I S]
600 (math-nth-root-int-iter a (or guess
601 (math-scale-int 1 (/ (+ (math-numdigs a)
602 (1- n))
603 n)))))
605 (defun math-nth-root-int-iter (a guess) ; uses "n"
606 (math-working "root" guess)
607 (let* ((q (math-idivmod a (math-ipow guess (1- n))))
608 (s (math-add (car q) (math-mul (1- n) guess)))
609 (g2 (math-idivmod s n)))
610 (if (Math-natnum-lessp (car g2) guess)
611 (math-nth-root-int-iter a (car g2))
612 (cons (and (equal (car g2) guess)
613 (eq (cdr q) 0)
614 (eq (cdr g2) 0))
615 guess))))
617 (defun calcFunc-nroot (x n)
618 (calcFunc-pow x (if (integerp n)
619 (math-make-frac 1 n)
620 (math-div 1 n))))
625 ;;;; Transcendental functions.
627 ;;; All of these functions are defined on the complex plane.
628 ;;; (Branch cuts, etc. follow Steele's Common Lisp book.)
630 ;;; Most functions increase calc-internal-prec by 2 digits, then round
631 ;;; down afterward. "-raw" functions use the current precision, require
632 ;;; their arguments to be in float (or complex float) format, and always
633 ;;; work in radians (where applicable).
635 (defun math-to-radians (a) ; [N N]
636 (cond ((eq (car-safe a) 'hms)
637 (math-from-hms a 'rad))
638 ((memq calc-angle-mode '(deg hms))
639 (math-mul a (math-pi-over-180)))
640 (t a)))
642 (defun math-from-radians (a) ; [N N]
643 (cond ((eq calc-angle-mode 'deg)
644 (if (math-constp a)
645 (math-div a (math-pi-over-180))
646 (list 'calcFunc-deg a)))
647 ((eq calc-angle-mode 'hms)
648 (math-to-hms a 'rad))
649 (t a)))
651 (defun math-to-radians-2 (a) ; [N N]
652 (cond ((eq (car-safe a) 'hms)
653 (math-from-hms a 'rad))
654 ((memq calc-angle-mode '(deg hms))
655 (if calc-symbolic-mode
656 (math-div (math-mul a '(var pi var-pi)) 180)
657 (math-mul a (math-pi-over-180))))
658 (t a)))
660 (defun math-from-radians-2 (a) ; [N N]
661 (cond ((memq calc-angle-mode '(deg hms))
662 (if calc-symbolic-mode
663 (math-div (math-mul 180 a) '(var pi var-pi))
664 (math-div a (math-pi-over-180))))
665 (t a)))
669 ;;; Sine, cosine, and tangent.
671 (defun calcFunc-sin (x) ; [N N] [Public]
672 (cond ((and (integerp x)
673 (if (eq calc-angle-mode 'deg)
674 (= (% x 90) 0)
675 (= x 0)))
676 (aref [0 1 0 -1] (math-mod (/ x 90) 4)))
677 ((Math-scalarp x)
678 (math-with-extra-prec 2
679 (math-sin-raw (math-to-radians (math-float x)))))
680 ((eq (car x) 'sdev)
681 (if (math-constp x)
682 (math-with-extra-prec 2
683 (let* ((xx (math-to-radians (math-float (nth 1 x))))
684 (xs (math-to-radians (math-float (nth 2 x))))
685 (sc (math-sin-cos-raw xx)))
686 (math-make-sdev (car sc) (math-mul xs (cdr sc)))))
687 (math-make-sdev (calcFunc-sin (nth 1 x))
688 (math-mul (nth 2 x) (calcFunc-cos (nth 1 x))))))
689 ((and (eq (car x) 'intv) (math-intv-constp x))
690 (calcFunc-cos (math-sub x (math-quarter-circle nil))))
691 ((equal x '(var nan var-nan))
693 (t (calc-record-why 'scalarp x)
694 (list 'calcFunc-sin x))))
696 (defun calcFunc-cos (x) ; [N N] [Public]
697 (cond ((and (integerp x)
698 (if (eq calc-angle-mode 'deg)
699 (= (% x 90) 0)
700 (= x 0)))
701 (aref [1 0 -1 0] (math-mod (/ x 90) 4)))
702 ((Math-scalarp x)
703 (math-with-extra-prec 2
704 (math-cos-raw (math-to-radians (math-float x)))))
705 ((eq (car x) 'sdev)
706 (if (math-constp x)
707 (math-with-extra-prec 2
708 (let* ((xx (math-to-radians (math-float (nth 1 x))))
709 (xs (math-to-radians (math-float (nth 2 x))))
710 (sc (math-sin-cos-raw xx)))
711 (math-make-sdev (cdr sc) (math-mul xs (car sc)))))
712 (math-make-sdev (calcFunc-cos (nth 1 x))
713 (math-mul (nth 2 x) (calcFunc-sin (nth 1 x))))))
714 ((and (eq (car x) 'intv) (math-intv-constp x))
715 (math-with-extra-prec 2
716 (let* ((xx (math-to-radians (math-float x)))
717 (na (math-floor (math-div (nth 2 xx) (math-pi))))
718 (nb (math-floor (math-div (nth 3 xx) (math-pi))))
719 (span (math-sub nb na)))
720 (if (memq span '(0 1))
721 (let ((int (math-sort-intv (nth 1 x)
722 (math-cos-raw (nth 2 xx))
723 (math-cos-raw (nth 3 xx)))))
724 (if (eq span 1)
725 (if (math-evenp na)
726 (math-make-intv (logior (nth 1 x) 2)
728 (nth 3 int))
729 (math-make-intv (logior (nth 1 x) 1)
730 (nth 2 int)
732 int))
733 (list 'intv 3 -1 1)))))
734 ((equal x '(var nan var-nan))
736 (t (calc-record-why 'scalarp x)
737 (list 'calcFunc-cos x))))
739 (defun calcFunc-sincos (x) ; [V N] [Public]
740 (if (Math-scalarp x)
741 (math-with-extra-prec 2
742 (let ((sc (math-sin-cos-raw (math-to-radians (math-float x)))))
743 (list 'vec (cdr sc) (car sc)))) ; the vector [cos, sin]
744 (list 'vec (calcFunc-sin x) (calcFunc-cos x))))
746 (defun calcFunc-tan (x) ; [N N] [Public]
747 (cond ((and (integerp x)
748 (if (eq calc-angle-mode 'deg)
749 (= (% x 180) 0)
750 (= x 0)))
752 ((Math-scalarp x)
753 (math-with-extra-prec 2
754 (math-tan-raw (math-to-radians (math-float x)))))
755 ((eq (car x) 'sdev)
756 (if (math-constp x)
757 (math-with-extra-prec 2
758 (let* ((xx (math-to-radians (math-float (nth 1 x))))
759 (xs (math-to-radians (math-float (nth 2 x))))
760 (sc (math-sin-cos-raw xx)))
761 (if (and (math-zerop (cdr sc)) (not calc-infinite-mode))
762 (progn
763 (calc-record-why "*Division by zero")
764 (list 'calcFunc-tan x))
765 (math-make-sdev (math-div-float (car sc) (cdr sc))
766 (math-div-float xs (math-sqr (cdr sc)))))))
767 (math-make-sdev (calcFunc-tan (nth 1 x))
768 (math-div (nth 2 x)
769 (math-sqr (calcFunc-cos (nth 1 x)))))))
770 ((and (eq (car x) 'intv) (math-intv-constp x))
771 (or (math-with-extra-prec 2
772 (let* ((xx (math-to-radians (math-float x)))
773 (na (math-floor (math-div (math-sub (nth 2 xx)
774 (math-pi-over-2))
775 (math-pi))))
776 (nb (math-floor (math-div (math-sub (nth 3 xx)
777 (math-pi-over-2))
778 (math-pi)))))
779 (and (equal na nb)
780 (math-sort-intv (nth 1 x)
781 (math-tan-raw (nth 2 xx))
782 (math-tan-raw (nth 3 xx))))))
783 '(intv 3 (neg (var inf var-inf)) (var inf var-inf))))
784 ((equal x '(var nan var-nan))
786 (t (calc-record-why 'scalarp x)
787 (list 'calcFunc-tan x))))
789 (defun math-sin-raw (x) ; [N N]
790 (cond ((eq (car x) 'cplx)
791 (let* ((expx (math-exp-raw (nth 2 x)))
792 (expmx (math-div-float '(float 1 0) expx))
793 (sc (math-sin-cos-raw (nth 1 x))))
794 (list 'cplx
795 (math-mul-float (car sc)
796 (math-mul-float (math-add-float expx expmx)
797 '(float 5 -1)))
798 (math-mul-float (cdr sc)
799 (math-mul-float (math-sub-float expx expmx)
800 '(float 5 -1))))))
801 ((eq (car x) 'polar)
802 (math-polar (math-sin-raw (math-complex x))))
803 ((Math-integer-negp (nth 1 x))
804 (math-neg-float (math-sin-raw (math-neg-float x))))
805 ((math-lessp-float '(float 7 0) x) ; avoid inf loops due to roundoff
806 (math-sin-raw (math-mod x (math-two-pi))))
807 (t (math-sin-raw-2 x x))))
809 (defun math-cos-raw (x) ; [N N]
810 (if (eq (car-safe x) 'polar)
811 (math-polar (math-cos-raw (math-complex x)))
812 (math-sin-raw (math-sub (math-pi-over-2) x))))
814 ;;; This could use a smarter method: Reduce x as in math-sin-raw, then
815 ;;; compute either sin(x) or cos(x), whichever is smaller, and compute
816 ;;; the other using the identity sin(x)^2 + cos(x)^2 = 1.
817 (defun math-sin-cos-raw (x) ; [F.F F] (result is (sin x . cos x))
818 (cons (math-sin-raw x) (math-cos-raw x)))
820 (defun math-tan-raw (x) ; [N N]
821 (cond ((eq (car x) 'cplx)
822 (let* ((x (math-mul x '(float 2 0)))
823 (expx (math-exp-raw (nth 2 x)))
824 (expmx (math-div-float '(float 1 0) expx))
825 (sc (math-sin-cos-raw (nth 1 x)))
826 (d (math-add-float (cdr sc)
827 (math-mul-float (math-add-float expx expmx)
828 '(float 5 -1)))))
829 (and (not (eq (nth 1 d) 0))
830 (list 'cplx
831 (math-div-float (car sc) d)
832 (math-div-float (math-mul-float (math-sub-float expx
833 expmx)
834 '(float 5 -1)) d)))))
835 ((eq (car x) 'polar)
836 (math-polar (math-tan-raw (math-complex x))))
838 (let ((sc (math-sin-cos-raw x)))
839 (if (eq (nth 1 (cdr sc)) 0)
840 (math-div (car sc) 0)
841 (math-div-float (car sc) (cdr sc)))))))
843 (defun math-sin-raw-2 (x orgx) ; This avoids poss of inf recursion. [F F]
844 (let ((xmpo2 (math-sub-float (math-pi-over-2) x)))
845 (cond ((Math-integer-negp (nth 1 xmpo2))
846 (math-neg-float (math-sin-raw-2 (math-sub-float x (math-pi))
847 orgx)))
848 ((math-lessp-float (math-pi-over-4) x)
849 (math-cos-raw-2 xmpo2 orgx))
850 ((math-lessp-float x (math-neg (math-pi-over-4)))
851 (math-neg (math-cos-raw-2 (math-add (math-pi-over-2) x) orgx)))
852 ((math-nearly-zerop-float x orgx) '(float 0 0))
853 (calc-symbolic-mode (signal 'inexact-result nil))
854 (t (math-sin-series x 6 4 x (math-neg-float (math-sqr-float x)))))))
856 (defun math-cos-raw-2 (x orgx) ; [F F]
857 (cond ((math-nearly-zerop-float x orgx) '(float 1 0))
858 (calc-symbolic-mode (signal 'inexact-result nil))
859 (t (let ((xnegsqr (math-neg-float (math-sqr-float x))))
860 (math-sin-series
861 (math-add-float '(float 1 0)
862 (math-mul-float xnegsqr '(float 5 -1)))
863 24 5 xnegsqr xnegsqr)))))
865 (defun math-sin-series (sum nfac n x xnegsqr)
866 (math-working "sin" sum)
867 (let* ((nextx (math-mul-float x xnegsqr))
868 (nextsum (math-add-float sum (math-div-float nextx
869 (math-float nfac)))))
870 (if (math-nearly-equal-float sum nextsum)
872 (math-sin-series nextsum (math-mul nfac (* n (1+ n)))
873 (+ n 2) nextx xnegsqr))))
876 ;;; Inverse sine, cosine, tangent.
878 (defun calcFunc-arcsin (x) ; [N N] [Public]
879 (cond ((eq x 0) 0)
880 ((and (eq x 1) (eq calc-angle-mode 'deg)) 90)
881 ((and (eq x -1) (eq calc-angle-mode 'deg)) -90)
882 (calc-symbolic-mode (signal 'inexact-result nil))
883 ((Math-numberp x)
884 (math-with-extra-prec 2
885 (math-from-radians (math-arcsin-raw (math-float x)))))
886 ((eq (car x) 'sdev)
887 (math-make-sdev (calcFunc-arcsin (nth 1 x))
888 (math-from-radians
889 (math-div (nth 2 x)
890 (math-sqrt
891 (math-sub 1 (math-sqr (nth 1 x))))))))
892 ((eq (car x) 'intv)
893 (math-sort-intv (nth 1 x)
894 (calcFunc-arcsin (nth 2 x))
895 (calcFunc-arcsin (nth 3 x))))
896 ((equal x '(var nan var-nan))
898 (t (calc-record-why 'numberp x)
899 (list 'calcFunc-arcsin x))))
901 (defun calcFunc-arccos (x) ; [N N] [Public]
902 (cond ((eq x 1) 0)
903 ((and (eq x 0) (eq calc-angle-mode 'deg)) 90)
904 ((and (eq x -1) (eq calc-angle-mode 'deg)) 180)
905 (calc-symbolic-mode (signal 'inexact-result nil))
906 ((Math-numberp x)
907 (math-with-extra-prec 2
908 (math-from-radians (math-arccos-raw (math-float x)))))
909 ((eq (car x) 'sdev)
910 (math-make-sdev (calcFunc-arccos (nth 1 x))
911 (math-from-radians
912 (math-div (nth 2 x)
913 (math-sqrt
914 (math-sub 1 (math-sqr (nth 1 x))))))))
915 ((eq (car x) 'intv)
916 (math-sort-intv (nth 1 x)
917 (calcFunc-arccos (nth 2 x))
918 (calcFunc-arccos (nth 3 x))))
919 ((equal x '(var nan var-nan))
921 (t (calc-record-why 'numberp x)
922 (list 'calcFunc-arccos x))))
924 (defun calcFunc-arctan (x) ; [N N] [Public]
925 (cond ((eq x 0) 0)
926 ((and (eq x 1) (eq calc-angle-mode 'deg)) 45)
927 ((and (eq x -1) (eq calc-angle-mode 'deg)) -45)
928 ((Math-numberp x)
929 (math-with-extra-prec 2
930 (math-from-radians (math-arctan-raw (math-float x)))))
931 ((eq (car x) 'sdev)
932 (math-make-sdev (calcFunc-arctan (nth 1 x))
933 (math-from-radians
934 (math-div (nth 2 x)
935 (math-add 1 (math-sqr (nth 1 x)))))))
936 ((eq (car x) 'intv)
937 (math-sort-intv (nth 1 x)
938 (calcFunc-arctan (nth 2 x))
939 (calcFunc-arctan (nth 3 x))))
940 ((equal x '(var inf var-inf))
941 (math-quarter-circle t))
942 ((equal x '(neg (var inf var-inf)))
943 (math-neg (math-quarter-circle t)))
944 ((equal x '(var nan var-nan))
946 (t (calc-record-why 'numberp x)
947 (list 'calcFunc-arctan x))))
949 (defun math-arcsin-raw (x) ; [N N]
950 (let ((a (math-sqrt-raw (math-sub '(float 1 0) (math-sqr x)))))
951 (if (or (memq (car x) '(cplx polar))
952 (memq (car a) '(cplx polar)))
953 (math-with-extra-prec 2 ; use extra precision for difficult case
954 (math-mul '(cplx 0 -1)
955 (math-ln-raw (math-add (math-mul '(cplx 0 1) x) a))))
956 (math-arctan2-raw x a))))
958 (defun math-arccos-raw (x) ; [N N]
959 (math-sub (math-pi-over-2) (math-arcsin-raw x)))
961 (defun math-arctan-raw (x) ; [N N]
962 (cond ((memq (car x) '(cplx polar))
963 (math-with-extra-prec 2 ; extra-extra
964 (math-div (math-sub
965 (math-ln-raw (math-add 1 (math-mul '(cplx 0 1) x)))
966 (math-ln-raw (math-add 1 (math-mul '(cplx 0 -1) x))))
967 '(cplx 0 2))))
968 ((Math-integer-negp (nth 1 x))
969 (math-neg-float (math-arctan-raw (math-neg-float x))))
970 ((math-zerop x) x)
971 (calc-symbolic-mode (signal 'inexact-result nil))
972 ((math-equal-int x 1) (math-pi-over-4))
973 ((math-equal-int x -1) (math-neg (math-pi-over-4)))
974 ((math-lessp-float '(float 414214 -6) x) ; if x > sqrt(2) - 1, reduce
975 (if (math-lessp-float '(float 1 0) x)
976 (math-sub-float (math-mul-float (math-pi) '(float 5 -1))
977 (math-arctan-raw (math-div-float '(float 1 0) x)))
978 (math-sub-float (math-mul-float (math-pi) '(float 25 -2))
979 (math-arctan-raw (math-div-float
980 (math-sub-float '(float 1 0) x)
981 (math-add-float '(float 1 0)
982 x))))))
983 (t (math-arctan-series x 3 x (math-neg-float (math-sqr-float x))))))
985 (defun math-arctan-series (sum n x xnegsqr)
986 (math-working "arctan" sum)
987 (let* ((nextx (math-mul-float x xnegsqr))
988 (nextsum (math-add-float sum (math-div-float nextx (math-float n)))))
989 (if (math-nearly-equal-float sum nextsum)
991 (math-arctan-series nextsum (+ n 2) nextx xnegsqr))))
993 (defun calcFunc-arctan2 (y x) ; [F R R] [Public]
994 (if (Math-anglep y)
995 (if (Math-anglep x)
996 (math-with-extra-prec 2
997 (math-from-radians (math-arctan2-raw (math-float y)
998 (math-float x))))
999 (calc-record-why 'anglep x)
1000 (list 'calcFunc-arctan2 y x))
1001 (if (and (or (math-infinitep x) (math-anglep x))
1002 (or (math-infinitep y) (math-anglep y)))
1003 (progn
1004 (if (math-posp x)
1005 (setq x 1)
1006 (if (math-negp x)
1007 (setq x -1)
1008 (or (math-zerop x)
1009 (setq x nil))))
1010 (if (math-posp y)
1011 (setq y 1)
1012 (if (math-negp y)
1013 (setq y -1)
1014 (or (math-zerop y)
1015 (setq y nil))))
1016 (if (and y x)
1017 (calcFunc-arctan2 y x)
1018 '(var nan var-nan)))
1019 (calc-record-why 'anglep y)
1020 (list 'calcFunc-arctan2 y x))))
1022 (defun math-arctan2-raw (y x) ; [F R R]
1023 (cond ((math-zerop y)
1024 (if (math-negp x) (math-pi)
1025 (if (or (math-floatp x) (math-floatp y)) '(float 0 0) 0)))
1026 ((math-zerop x)
1027 (if (math-posp y)
1028 (math-pi-over-2)
1029 (math-neg (math-pi-over-2))))
1030 ((math-posp x)
1031 (math-arctan-raw (math-div-float y x)))
1032 ((math-posp y)
1033 (math-add-float (math-arctan-raw (math-div-float y x))
1034 (math-pi)))
1036 (math-sub-float (math-arctan-raw (math-div-float y x))
1037 (math-pi)))))
1039 (defun calcFunc-arcsincos (x) ; [V N] [Public]
1040 (if (and (Math-vectorp x)
1041 (= (length x) 3))
1042 (calcFunc-arctan2 (nth 2 x) (nth 1 x))
1043 (math-reject-arg x "*Two-element vector expected")))
1047 ;;; Exponential function.
1049 (defun calcFunc-exp (x) ; [N N] [Public]
1050 (cond ((eq x 0) 1)
1051 ((and (memq x '(1 -1)) calc-symbolic-mode)
1052 (if (eq x 1) '(var e var-e) (math-div 1 '(var e var-e))))
1053 ((Math-numberp x)
1054 (math-with-extra-prec 2 (math-exp-raw (math-float x))))
1055 ((eq (car-safe x) 'sdev)
1056 (let ((ex (calcFunc-exp (nth 1 x))))
1057 (math-make-sdev ex (math-mul (nth 2 x) ex))))
1058 ((eq (car-safe x) 'intv)
1059 (math-make-intv (nth 1 x) (calcFunc-exp (nth 2 x))
1060 (calcFunc-exp (nth 3 x))))
1061 ((equal x '(var inf var-inf))
1063 ((equal x '(neg (var inf var-inf)))
1065 ((equal x '(var nan var-nan))
1067 (t (calc-record-why 'numberp x)
1068 (list 'calcFunc-exp x))))
1070 (defun calcFunc-expm1 (x) ; [N N] [Public]
1071 (cond ((eq x 0) 0)
1072 ((math-zerop x) '(float 0 0))
1073 (calc-symbolic-mode (signal 'inexact-result nil))
1074 ((Math-numberp x)
1075 (math-with-extra-prec 2
1076 (let ((x (math-float x)))
1077 (if (and (eq (car x) 'float)
1078 (math-lessp-float x '(float 1 0))
1079 (math-lessp-float '(float -1 0) x))
1080 (math-exp-minus-1-raw x)
1081 (math-add (math-exp-raw x) -1)))))
1082 ((eq (car-safe x) 'sdev)
1083 (if (math-constp x)
1084 (let ((ex (calcFunc-expm1 (nth 1 x))))
1085 (math-make-sdev ex (math-mul (nth 2 x) (math-add ex 1))))
1086 (math-make-sdev (calcFunc-expm1 (nth 1 x))
1087 (math-mul (nth 2 x) (calcFunc-exp (nth 1 x))))))
1088 ((eq (car-safe x) 'intv)
1089 (math-make-intv (nth 1 x)
1090 (calcFunc-expm1 (nth 2 x))
1091 (calcFunc-expm1 (nth 3 x))))
1092 ((equal x '(var inf var-inf))
1094 ((equal x '(neg (var inf var-inf)))
1096 ((equal x '(var nan var-nan))
1098 (t (calc-record-why 'numberp x)
1099 (list 'calcFunc-expm1 x))))
1101 (defun calcFunc-exp10 (x) ; [N N] [Public]
1102 (if (eq x 0)
1104 (math-pow '(float 1 1) x)))
1106 (defun math-exp-raw (x) ; [N N]
1107 (cond ((math-zerop x) '(float 1 0))
1108 (calc-symbolic-mode (signal 'inexact-result nil))
1109 ((eq (car x) 'cplx)
1110 (let ((expx (math-exp-raw (nth 1 x)))
1111 (sc (math-sin-cos-raw (nth 2 x))))
1112 (list 'cplx
1113 (math-mul-float expx (cdr sc))
1114 (math-mul-float expx (car sc)))))
1115 ((eq (car x) 'polar)
1116 (let ((xc (math-complex x)))
1117 (list 'polar
1118 (math-exp-raw (nth 1 xc))
1119 (math-from-radians (nth 2 xc)))))
1120 ((or (math-lessp-float '(float 5 -1) x)
1121 (math-lessp-float x '(float -5 -1)))
1122 (if (math-lessp-float '(float 921035 1) x)
1123 (math-overflow)
1124 (if (math-lessp-float x '(float -921035 1))
1125 (math-underflow)))
1126 (let* ((two-x (math-mul-float x '(float 2 0)))
1127 (hint (math-scale-int (nth 1 two-x) (nth 2 two-x)))
1128 (hfrac (math-sub-float x (math-mul-float (math-float hint)
1129 '(float 5 -1)))))
1130 (math-mul-float (math-ipow (math-sqrt-e) hint)
1131 (math-add-float '(float 1 0)
1132 (math-exp-minus-1-raw hfrac)))))
1133 (t (math-add-float '(float 1 0) (math-exp-minus-1-raw x)))))
1135 (defun math-exp-minus-1-raw (x) ; [F F]
1136 (math-exp-series x 2 3 x x))
1138 (defun math-exp-series (sum nfac n xpow x)
1139 (math-working "exp" sum)
1140 (let* ((nextx (math-mul-float xpow x))
1141 (nextsum (math-add-float sum (math-div-float nextx
1142 (math-float nfac)))))
1143 (if (math-nearly-equal-float sum nextsum)
1145 (math-exp-series nextsum (math-mul nfac n) (1+ n) nextx x))))
1149 ;;; Logarithms.
1151 (defun calcFunc-ln (x) ; [N N] [Public]
1152 (cond ((math-zerop x)
1153 (if calc-infinite-mode
1154 '(neg (var inf var-inf))
1155 (math-reject-arg x "*Logarithm of zero")))
1156 ((eq x 1) 0)
1157 ((Math-numberp x)
1158 (math-with-extra-prec 2 (math-ln-raw (math-float x))))
1159 ((eq (car-safe x) 'sdev)
1160 (math-make-sdev (calcFunc-ln (nth 1 x))
1161 (math-div (nth 2 x) (nth 1 x))))
1162 ((and (eq (car-safe x) 'intv) (or (Math-posp (nth 2 x))
1163 (Math-zerop (nth 2 x))
1164 (not (math-intv-constp x))))
1165 (let ((calc-infinite-mode t))
1166 (math-make-intv (nth 1 x) (calcFunc-ln (nth 2 x))
1167 (calcFunc-ln (nth 3 x)))))
1168 ((equal x '(var e var-e))
1170 ((and (eq (car-safe x) '^)
1171 (equal (nth 1 x) '(var e var-e))
1172 (math-known-realp (nth 2 x)))
1173 (nth 2 x))
1174 ((math-infinitep x)
1175 (if (equal x '(var nan var-nan))
1177 '(var inf var-inf)))
1178 (t (calc-record-why 'numberp x)
1179 (list 'calcFunc-ln x))))
1181 (defun calcFunc-log10 (x) ; [N N] [Public]
1182 (cond ((math-equal-int x 1)
1183 (if (math-floatp x) '(float 0 0) 0))
1184 ((and (Math-integerp x)
1185 (math-posp x)
1186 (let ((res (math-integer-log x 10)))
1187 (and (car res)
1188 (setq x (cdr res)))))
1190 ((and (eq (car-safe x) 'frac)
1191 (eq (nth 1 x) 1)
1192 (let ((res (math-integer-log (nth 2 x) 10)))
1193 (and (car res)
1194 (setq x (- (cdr res))))))
1196 ((math-zerop x)
1197 (if calc-infinite-mode
1198 '(neg (var inf var-inf))
1199 (math-reject-arg x "*Logarithm of zero")))
1200 (calc-symbolic-mode (signal 'inexact-result nil))
1201 ((Math-numberp x)
1202 (math-with-extra-prec 2
1203 (let ((xf (math-float x)))
1204 (if (eq (nth 1 xf) 0)
1205 (math-reject-arg x "*Logarithm of zero"))
1206 (if (Math-integer-posp (nth 1 xf))
1207 (if (eq (nth 1 xf) 1) ; log10(1*10^n) = n
1208 (math-float (nth 2 xf))
1209 (let ((xdigs (1- (math-numdigs (nth 1 xf)))))
1210 (math-add-float
1211 (math-div-float (math-ln-raw-2
1212 (list 'float (nth 1 xf) (- xdigs)))
1213 (math-ln-10))
1214 (math-float (+ (nth 2 xf) xdigs)))))
1215 (math-div (calcFunc-ln xf) (math-ln-10))))))
1216 ((eq (car-safe x) 'sdev)
1217 (math-make-sdev (calcFunc-log10 (nth 1 x))
1218 (math-div (nth 2 x)
1219 (math-mul (nth 1 x) (math-ln-10)))))
1220 ((and (eq (car-safe x) 'intv) (or (Math-posp (nth 2 x))
1221 (not (math-intv-constp x))))
1222 (math-make-intv (nth 1 x)
1223 (calcFunc-log10 (nth 2 x))
1224 (calcFunc-log10 (nth 3 x))))
1225 ((math-infinitep x)
1226 (if (equal x '(var nan var-nan))
1228 '(var inf var-inf)))
1229 (t (calc-record-why 'numberp x)
1230 (list 'calcFunc-log10 x))))
1232 (defun calcFunc-log (x &optional b) ; [N N N] [Public]
1233 (cond ((or (null b) (equal b '(var e var-e)))
1234 (math-normalize (list 'calcFunc-ln x)))
1235 ((or (eq b 10) (equal b '(float 1 1)))
1236 (math-normalize (list 'calcFunc-log10 x)))
1237 ((math-zerop x)
1238 (if calc-infinite-mode
1239 (math-div (calcFunc-ln x) (calcFunc-ln b))
1240 (math-reject-arg x "*Logarithm of zero")))
1241 ((math-zerop b)
1242 (if calc-infinite-mode
1243 (math-div (calcFunc-ln x) (calcFunc-ln b))
1244 (math-reject-arg b "*Logarithm of zero")))
1245 ((math-equal-int b 1)
1246 (if calc-infinite-mode
1247 (math-div (calcFunc-ln x) 0)
1248 (math-reject-arg b "*Logarithm base one")))
1249 ((math-equal-int x 1)
1250 (if (or (math-floatp a) (math-floatp b)) '(float 0 0) 0))
1251 ((and (Math-ratp x) (Math-ratp b)
1252 (math-posp x) (math-posp b)
1253 (let* ((sign 1) (inv nil)
1254 (xx (if (Math-lessp 1 x)
1256 (setq sign -1)
1257 (math-div 1 x)))
1258 (bb (if (Math-lessp 1 b)
1260 (setq sign (- sign))
1261 (math-div 1 b)))
1262 (res (if (Math-lessp xx bb)
1263 (setq inv (math-integer-log bb xx))
1264 (math-integer-log xx bb))))
1265 (and (car res)
1266 (setq x (if inv
1267 (math-div 1 (* sign (cdr res)))
1268 (* sign (cdr res)))))))
1270 (calc-symbolic-mode (signal 'inexact-result nil))
1271 ((and (Math-numberp x) (Math-numberp b))
1272 (math-with-extra-prec 2
1273 (math-div (math-ln-raw (math-float x))
1274 (math-log-base-raw b))))
1275 ((and (eq (car-safe x) 'sdev)
1276 (Math-numberp b))
1277 (math-make-sdev (calcFunc-log (nth 1 x) b)
1278 (math-div (nth 2 x)
1279 (math-mul (nth 1 x)
1280 (math-log-base-raw b)))))
1281 ((and (eq (car-safe x) 'intv) (or (Math-posp (nth 2 x))
1282 (not (math-intv-constp x)))
1283 (math-realp b))
1284 (math-make-intv (nth 1 x)
1285 (calcFunc-log (nth 2 x) b)
1286 (calcFunc-log (nth 3 x) b)))
1287 ((or (eq (car-safe x) 'intv) (eq (car-safe b) 'intv))
1288 (math-div (calcFunc-ln x) (calcFunc-ln b)))
1289 ((or (math-infinitep x)
1290 (math-infinitep b))
1291 (math-div (calcFunc-ln x) (calcFunc-ln b)))
1292 (t (if (Math-numberp b)
1293 (calc-record-why 'numberp x)
1294 (calc-record-why 'numberp b))
1295 (list 'calcFunc-log x b))))
1297 (defun calcFunc-alog (x &optional b)
1298 (cond ((or (null b) (equal b '(var e var-e)))
1299 (math-normalize (list 'calcFunc-exp x)))
1300 (t (math-pow b x))))
1302 (defun calcFunc-ilog (x b)
1303 (if (and (math-natnump x) (not (eq x 0))
1304 (math-natnump b) (not (eq b 0)))
1305 (if (eq b 1)
1306 (math-reject-arg x "*Logarithm base one")
1307 (if (Math-natnum-lessp x b)
1309 (cdr (math-integer-log x b))))
1310 (math-floor (calcFunc-log x b))))
1312 (defun math-integer-log (x b)
1313 (let ((pows (list b))
1314 (pow (math-sqr b))
1315 next
1316 sum n)
1317 (while (not (Math-lessp x pow))
1318 (setq pows (cons pow pows)
1319 pow (math-sqr pow)))
1320 (setq n (lsh 1 (1- (length pows)))
1321 sum n
1322 pow (car pows))
1323 (while (and (setq pows (cdr pows))
1324 (Math-lessp pow x))
1325 (setq n (/ n 2)
1326 next (math-mul pow (car pows)))
1327 (or (Math-lessp x next)
1328 (setq pow next
1329 sum (+ sum n))))
1330 (cons (equal pow x) sum)))
1333 (defvar math-log-base-cache nil)
1334 (defun math-log-base-raw (b) ; [N N]
1335 (if (not (and (equal (car math-log-base-cache) b)
1336 (eq (nth 1 math-log-base-cache) calc-internal-prec)))
1337 (setq math-log-base-cache (list b calc-internal-prec
1338 (math-ln-raw (math-float b)))))
1339 (nth 2 math-log-base-cache))
1341 (defun calcFunc-lnp1 (x) ; [N N] [Public]
1342 (cond ((Math-equal-int x -1)
1343 (if calc-infinite-mode
1344 '(neg (var inf var-inf))
1345 (math-reject-arg x "*Logarithm of zero")))
1346 ((eq x 0) 0)
1347 ((math-zerop x) '(float 0 0))
1348 (calc-symbolic-mode (signal 'inexact-result nil))
1349 ((Math-numberp x)
1350 (math-with-extra-prec 2
1351 (let ((x (math-float x)))
1352 (if (and (eq (car x) 'float)
1353 (math-lessp-float x '(float 5 -1))
1354 (math-lessp-float '(float -5 -1) x))
1355 (math-ln-plus-1-raw x)
1356 (math-ln-raw (math-add-float x '(float 1 0)))))))
1357 ((eq (car-safe x) 'sdev)
1358 (math-make-sdev (calcFunc-lnp1 (nth 1 x))
1359 (math-div (nth 2 x) (math-add (nth 1 x) 1))))
1360 ((and (eq (car-safe x) 'intv) (or (Math-posp (nth 2 x))
1361 (not (math-intv-constp x))))
1362 (math-make-intv (nth 1 x)
1363 (calcFunc-lnp1 (nth 2 x))
1364 (calcFunc-lnp1 (nth 3 x))))
1365 ((math-infinitep x)
1366 (if (equal x '(var nan var-nan))
1368 '(var inf var-inf)))
1369 (t (calc-record-why 'numberp x)
1370 (list 'calcFunc-lnp1 x))))
1372 (defun math-ln-raw (x) ; [N N] --- must be float format!
1373 (cond ((eq (car-safe x) 'cplx)
1374 (list 'cplx
1375 (math-mul-float (math-ln-raw
1376 (math-add-float (math-sqr-float (nth 1 x))
1377 (math-sqr-float (nth 2 x))))
1378 '(float 5 -1))
1379 (math-arctan2-raw (nth 2 x) (nth 1 x))))
1380 ((eq (car x) 'polar)
1381 (math-polar (list 'cplx
1382 (math-ln-raw (nth 1 x))
1383 (math-to-radians (nth 2 x)))))
1384 ((Math-equal-int x 1)
1385 '(float 0 0))
1386 (calc-symbolic-mode (signal 'inexact-result nil))
1387 ((math-posp (nth 1 x)) ; positive and real
1388 (let ((xdigs (1- (math-numdigs (nth 1 x)))))
1389 (math-add-float (math-ln-raw-2 (list 'float (nth 1 x) (- xdigs)))
1390 (math-mul-float (math-float (+ (nth 2 x) xdigs))
1391 (math-ln-10)))))
1392 ((math-zerop x)
1393 (math-reject-arg x "*Logarithm of zero"))
1394 ((eq calc-complex-mode 'polar) ; negative and real
1395 (math-polar
1396 (list 'cplx ; negative and real
1397 (math-ln-raw (math-neg-float x))
1398 (math-pi))))
1399 (t (list 'cplx ; negative and real
1400 (math-ln-raw (math-neg-float x))
1401 (math-pi)))))
1403 (defun math-ln-raw-2 (x) ; [F F]
1404 (cond ((math-lessp-float '(float 14 -1) x)
1405 (math-add-float (math-ln-raw-2 (math-mul-float x '(float 5 -1)))
1406 (math-ln-2)))
1407 (t ; now .7 < x <= 1.4
1408 (math-ln-raw-3 (math-div-float (math-sub-float x '(float 1 0))
1409 (math-add-float x '(float 1 0)))))))
1411 (defun math-ln-raw-3 (x) ; [F F]
1412 (math-mul-float (math-ln-raw-series x 3 x (math-sqr-float x))
1413 '(float 2 0)))
1415 ;;; Compute ln((1+x)/(1-x))
1416 (defun math-ln-raw-series (sum n x xsqr)
1417 (math-working "log" sum)
1418 (let* ((nextx (math-mul-float x xsqr))
1419 (nextsum (math-add-float sum (math-div-float nextx (math-float n)))))
1420 (if (math-nearly-equal-float sum nextsum)
1422 (math-ln-raw-series nextsum (+ n 2) nextx xsqr))))
1424 (defun math-ln-plus-1-raw (x)
1425 (math-lnp1-series x 2 x (math-neg x)))
1427 (defun math-lnp1-series (sum n xpow x)
1428 (math-working "lnp1" sum)
1429 (let* ((nextx (math-mul-float xpow x))
1430 (nextsum (math-add-float sum (math-div-float nextx (math-float n)))))
1431 (if (math-nearly-equal-float sum nextsum)
1433 (math-lnp1-series nextsum (1+ n) nextx x))))
1435 (math-defcache math-ln-10 (float (bigpos 018 684 045 994 092 585 302 2) -21)
1436 (math-ln-raw-2 '(float 1 1)))
1438 (math-defcache math-ln-2 (float (bigpos 417 309 945 559 180 147 693) -21)
1439 (math-ln-raw-3 (math-float '(frac 1 3))))
1443 ;;; Hyperbolic functions.
1445 (defun calcFunc-sinh (x) ; [N N] [Public]
1446 (cond ((eq x 0) 0)
1447 (math-expand-formulas
1448 (math-normalize
1449 (list '/ (list '- (list 'calcFunc-exp x)
1450 (list 'calcFunc-exp (list 'neg x))) 2)))
1451 ((Math-numberp x)
1452 (if calc-symbolic-mode (signal 'inexact-result nil))
1453 (math-with-extra-prec 2
1454 (let ((expx (math-exp-raw (math-float x))))
1455 (math-mul (math-add expx (math-div -1 expx)) '(float 5 -1)))))
1456 ((eq (car-safe x) 'sdev)
1457 (math-make-sdev (calcFunc-sinh (nth 1 x))
1458 (math-mul (nth 2 x) (calcFunc-cosh (nth 1 x)))))
1459 ((eq (car x) 'intv)
1460 (math-sort-intv (nth 1 x)
1461 (calcFunc-sinh (nth 2 x))
1462 (calcFunc-sinh (nth 3 x))))
1463 ((or (equal x '(var inf var-inf))
1464 (equal x '(neg (var inf var-inf)))
1465 (equal x '(var nan var-nan)))
1467 (t (calc-record-why 'numberp x)
1468 (list 'calcFunc-sinh x))))
1469 (put 'calcFunc-sinh 'math-expandable t)
1471 (defun calcFunc-cosh (x) ; [N N] [Public]
1472 (cond ((eq x 0) 1)
1473 (math-expand-formulas
1474 (math-normalize
1475 (list '/ (list '+ (list 'calcFunc-exp x)
1476 (list 'calcFunc-exp (list 'neg x))) 2)))
1477 ((Math-numberp x)
1478 (if calc-symbolic-mode (signal 'inexact-result nil))
1479 (math-with-extra-prec 2
1480 (let ((expx (math-exp-raw (math-float x))))
1481 (math-mul (math-add expx (math-div 1 expx)) '(float 5 -1)))))
1482 ((eq (car-safe x) 'sdev)
1483 (math-make-sdev (calcFunc-cosh (nth 1 x))
1484 (math-mul (nth 2 x)
1485 (calcFunc-sinh (nth 1 x)))))
1486 ((and (eq (car x) 'intv) (math-intv-constp x))
1487 (setq x (math-abs x))
1488 (math-sort-intv (nth 1 x)
1489 (calcFunc-cosh (nth 2 x))
1490 (calcFunc-cosh (nth 3 x))))
1491 ((or (equal x '(var inf var-inf))
1492 (equal x '(neg (var inf var-inf)))
1493 (equal x '(var nan var-nan)))
1494 (math-abs x))
1495 (t (calc-record-why 'numberp x)
1496 (list 'calcFunc-cosh x))))
1497 (put 'calcFunc-cosh 'math-expandable t)
1499 (defun calcFunc-tanh (x) ; [N N] [Public]
1500 (cond ((eq x 0) 0)
1501 (math-expand-formulas
1502 (math-normalize
1503 (let ((expx (list 'calcFunc-exp x))
1504 (expmx (list 'calcFunc-exp (list 'neg x))))
1505 (math-normalize
1506 (list '/ (list '- expx expmx) (list '+ expx expmx))))))
1507 ((Math-numberp x)
1508 (if calc-symbolic-mode (signal 'inexact-result nil))
1509 (math-with-extra-prec 2
1510 (let* ((expx (calcFunc-exp (math-float x)))
1511 (expmx (math-div 1 expx)))
1512 (math-div (math-sub expx expmx)
1513 (math-add expx expmx)))))
1514 ((eq (car-safe x) 'sdev)
1515 (math-make-sdev (calcFunc-tanh (nth 1 x))
1516 (math-div (nth 2 x)
1517 (math-sqr (calcFunc-cosh (nth 1 x))))))
1518 ((eq (car x) 'intv)
1519 (math-sort-intv (nth 1 x)
1520 (calcFunc-tanh (nth 2 x))
1521 (calcFunc-tanh (nth 3 x))))
1522 ((equal x '(var inf var-inf))
1524 ((equal x '(neg (var inf var-inf)))
1526 ((equal x '(var nan var-nan))
1528 (t (calc-record-why 'numberp x)
1529 (list 'calcFunc-tanh x))))
1530 (put 'calcFunc-tanh 'math-expandable t)
1532 (defun calcFunc-arcsinh (x) ; [N N] [Public]
1533 (cond ((eq x 0) 0)
1534 (math-expand-formulas
1535 (math-normalize
1536 (list 'calcFunc-ln (list '+ x (list 'calcFunc-sqrt
1537 (list '+ (list '^ x 2) 1))))))
1538 ((Math-numberp x)
1539 (if calc-symbolic-mode (signal 'inexact-result nil))
1540 (math-with-extra-prec 2
1541 (math-ln-raw (math-add x (math-sqrt-raw (math-add (math-sqr x)
1542 '(float 1 0)))))))
1543 ((eq (car-safe x) 'sdev)
1544 (math-make-sdev (calcFunc-arcsinh (nth 1 x))
1545 (math-div (nth 2 x)
1546 (math-sqrt
1547 (math-add (math-sqr (nth 1 x)) 1)))))
1548 ((eq (car x) 'intv)
1549 (math-sort-intv (nth 1 x)
1550 (calcFunc-arcsinh (nth 2 x))
1551 (calcFunc-arcsinh (nth 3 x))))
1552 ((or (equal x '(var inf var-inf))
1553 (equal x '(neg (var inf var-inf)))
1554 (equal x '(var nan var-nan)))
1556 (t (calc-record-why 'numberp x)
1557 (list 'calcFunc-arcsinh x))))
1558 (put 'calcFunc-arcsinh 'math-expandable t)
1560 (defun calcFunc-arccosh (x) ; [N N] [Public]
1561 (cond ((eq x 1) 0)
1562 ((and (eq x -1) calc-symbolic-mode)
1563 '(var pi var-pi))
1564 ((and (eq x 0) calc-symbolic-mode)
1565 (math-div (math-mul '(var pi var-pi) '(var i var-i)) 2))
1566 (math-expand-formulas
1567 (math-normalize
1568 (list 'calcFunc-ln (list '+ x (list 'calcFunc-sqrt
1569 (list '- (list '^ x 2) 1))))))
1570 ((Math-numberp x)
1571 (if calc-symbolic-mode (signal 'inexact-result nil))
1572 (if (Math-equal-int x -1)
1573 (math-imaginary (math-pi))
1574 (math-with-extra-prec 2
1575 (if (or t ; need to do this even in the real case!
1576 (memq (car-safe x) '(cplx polar)))
1577 (let ((xp1 (math-add 1 x))) ; this gets the branch cuts right
1578 (math-ln-raw
1579 (math-add x (math-mul xp1
1580 (math-sqrt-raw
1581 (math-div (math-sub
1583 '(float 1 0))
1584 xp1))))))
1585 (math-ln-raw
1586 (math-add x (math-sqrt-raw (math-add (math-sqr x)
1587 '(float -1 0)))))))))
1588 ((eq (car-safe x) 'sdev)
1589 (math-make-sdev (calcFunc-arccosh (nth 1 x))
1590 (math-div (nth 2 x)
1591 (math-sqrt
1592 (math-add (math-sqr (nth 1 x)) -1)))))
1593 ((eq (car x) 'intv)
1594 (math-sort-intv (nth 1 x)
1595 (calcFunc-arccosh (nth 2 x))
1596 (calcFunc-arccosh (nth 3 x))))
1597 ((or (equal x '(var inf var-inf))
1598 (equal x '(neg (var inf var-inf)))
1599 (equal x '(var nan var-nan)))
1601 (t (calc-record-why 'numberp x)
1602 (list 'calcFunc-arccosh x))))
1603 (put 'calcFunc-arccosh 'math-expandable t)
1605 (defun calcFunc-arctanh (x) ; [N N] [Public]
1606 (cond ((eq x 0) 0)
1607 ((and (Math-equal-int x 1) calc-infinite-mode)
1608 '(var inf var-inf))
1609 ((and (Math-equal-int x -1) calc-infinite-mode)
1610 '(neg (var inf var-inf)))
1611 (math-expand-formulas
1612 (list '/ (list '-
1613 (list 'calcFunc-ln (list '+ 1 x))
1614 (list 'calcFunc-ln (list '- 1 x))) 2))
1615 ((Math-numberp x)
1616 (if calc-symbolic-mode (signal 'inexact-result nil))
1617 (math-with-extra-prec 2
1618 (if (or (memq (car-safe x) '(cplx polar))
1619 (Math-lessp 1 x))
1620 (math-mul (math-sub (math-ln-raw (math-add '(float 1 0) x))
1621 (math-ln-raw (math-sub '(float 1 0) x)))
1622 '(float 5 -1))
1623 (if (and (math-equal-int x 1) calc-infinite-mode)
1624 '(var inf var-inf)
1625 (if (and (math-equal-int x -1) calc-infinite-mode)
1626 '(neg (var inf var-inf))
1627 (math-mul (math-ln-raw (math-div (math-add '(float 1 0) x)
1628 (math-sub 1 x)))
1629 '(float 5 -1)))))))
1630 ((eq (car-safe x) 'sdev)
1631 (math-make-sdev (calcFunc-arctanh (nth 1 x))
1632 (math-div (nth 2 x)
1633 (math-sub 1 (math-sqr (nth 1 x))))))
1634 ((eq (car x) 'intv)
1635 (math-sort-intv (nth 1 x)
1636 (calcFunc-arctanh (nth 2 x))
1637 (calcFunc-arctanh (nth 3 x))))
1638 ((equal x '(var nan var-nan))
1640 (t (calc-record-why 'numberp x)
1641 (list 'calcFunc-arctanh x))))
1642 (put 'calcFunc-arctanh 'math-expandable t)
1645 ;;; Convert A from HMS or degrees to radians.
1646 (defun calcFunc-rad (a) ; [R R] [Public]
1647 (cond ((or (Math-numberp a)
1648 (eq (car a) 'intv))
1649 (math-with-extra-prec 2
1650 (math-mul a (math-pi-over-180))))
1651 ((eq (car a) 'hms)
1652 (math-from-hms a 'rad))
1653 ((eq (car a) 'sdev)
1654 (math-make-sdev (calcFunc-rad (nth 1 a))
1655 (calcFunc-rad (nth 2 a))))
1656 (math-expand-formulas
1657 (math-div (math-mul a '(var pi var-pi)) 180))
1658 ((math-infinitep a) a)
1659 (t (list 'calcFunc-rad a))))
1660 (put 'calcFunc-rad 'math-expandable t)
1662 ;;; Convert A from HMS or radians to degrees.
1663 (defun calcFunc-deg (a) ; [R R] [Public]
1664 (cond ((or (Math-numberp a)
1665 (eq (car a) 'intv))
1666 (math-with-extra-prec 2
1667 (math-div a (math-pi-over-180))))
1668 ((eq (car a) 'hms)
1669 (math-from-hms a 'deg))
1670 ((eq (car a) 'sdev)
1671 (math-make-sdev (calcFunc-deg (nth 1 a))
1672 (calcFunc-deg (nth 2 a))))
1673 (math-expand-formulas
1674 (math-div (math-mul 180 a) '(var pi var-pi)))
1675 ((math-infinitep a) a)
1676 (t (list 'calcFunc-deg a))))
1677 (put 'calcFunc-deg 'math-expandable t)
1679 ;;; calc-math.el ends here