File:  [gforth] / gforth / float.fs
Revision 1.49: download - view: text, annotated - select for diffs
Thu Dec 28 14:52:20 2006 UTC (12 years, 9 months ago) by anton
Branches: MAIN
CVS tags: HEAD
added >OUTFILE ... OUTFILE<, >INFILE...INFILE< and use it in ~~

    1: \ High level floating point                            14jan94py
    2: 
    3: \ Copyright (C) 1995,1997,2003,2004,2005 Free Software Foundation, Inc.
    4: 
    5: \ This file is part of Gforth.
    6: 
    7: \ Gforth is free software; you can redistribute it and/or
    8: \ modify it under the terms of the GNU General Public License
    9: \ as published by the Free Software Foundation; either version 2
   10: \ of the License, or (at your option) any later version.
   11: 
   12: \ This program is distributed in the hope that it will be useful,
   13: \ but WITHOUT ANY WARRANTY; without even the implied warranty of
   14: \ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   15: \ GNU General Public License for more details.
   16: 
   17: \ You should have received a copy of the GNU General Public License
   18: \ along with this program; if not, write to the Free Software
   19: \ Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111, USA.
   20: 
   21: \ 1 cells 4 = [IF]
   22: \ ' cells   Alias sfloats
   23: \ ' cell+   Alias sfloat+
   24: \ ' align   Alias sfalign
   25: \ ' aligned Alias sfaligned
   26: \ [ELSE]
   27: \ : sfloats  2* 2* ;
   28: \ : sfloat+  4 + ;
   29: \ : sfaligned ( addr -- addr' )  3 + -4 and ;
   30: \ : sfalign ( -- )  here dup sfaligned swap ?DO  bl c,  LOOP ;
   31: \ [THEN]
   32: 
   33: \ 1 floats 8 = [IF]
   34: \ ' floats   Alias dfloats
   35: \ ' float+   Alias dfloat+
   36: \ ' falign   Alias dfalign
   37: \ ' faligned Alias dfaligned
   38: \ [ELSE]
   39: \ : dfloats  2* 2* 2* ;
   40: \ : dfloat+  8 + ;
   41: \ : dfaligned ( addr -- addr' )  7 + -8 and ;
   42: \ : dfalign ( -- )  here dup dfaligned swap ?DO  bl c,  LOOP ;
   43: \ [THEN]
   44: 
   45: : sfalign ( -- ) \ float-ext s-f-align
   46:     \G If the data-space pointer is not single-float-aligned, reserve
   47:     \G enough space to align it.
   48:     here dup sfaligned swap ?DO  bl c,  LOOP ;
   49: : dfalign ( -- ) \ float-ext d-f-align
   50:     \G If the data-space pointer is not double-float-aligned, reserve
   51:     \G enough space to align it.
   52:     here dup dfaligned swap ?DO  bl c,  LOOP ;
   53: 
   54: 1 sfloats (Field) sfloat+ , ( sf-addr1 -- sf-addr2 ) \ float-ext s-float-plus
   55: \G @code{1 sfloats +}.
   56: 
   57: 1 dfloats (Field) dfloat+ , ( df-addr1 -- df-addr2 ) \ float-ext d-float-plus
   58: \G @code{1 dfloats +}.
   59: 
   60: : f, ( f -- ) \ gforth
   61:     \G Reserve data space for one floating-point number and store
   62:     \G @i{f} in the space.
   63:     here 1 floats allot f! ;
   64: 
   65: : fconstant  ( r "name" -- ) \ float f-constant
   66:     Create f,
   67: DOES> ( -- r )
   68:     f@ ;
   69: 
   70: : fdepth ( -- +n ) \ float f-depth
   71:     \G @i{+n} is the current number of (floating-point) values on the
   72:     \G floating-point stack.
   73:     fp0 @ fp@ - [ 1 floats ] Literal / ;
   74: 
   75: : FLiteral ( compilation r -- ; run-time -- r ) \ float f-literal
   76:     \G Compile appropriate code such that, at run-time, @i{r} is placed
   77:     \G on the (floating-point) stack. Interpretation semantics are undefined.
   78:     BEGIN  here cell+ cell+ dup faligned <>  WHILE  postpone noop  REPEAT
   79:     postpone ahead here >r f, postpone then
   80:     r> postpone literal postpone f@ ;  immediate
   81: 
   82: &15 Value precision ( -- u ) \ float-ext
   83: \G @i{u} is the number of significant digits currently used by
   84: \G @code{F.} @code{FE.} and @code{FS.} 
   85: : set-precision ( u -- ) \ float-ext
   86:     \G Set the number of significant digits currently used by
   87:     \G @code{F.} @code{FE.} and @code{FS.} to @i{u}.
   88:     to precision ;
   89: 
   90: : scratch ( r -- addr len )
   91:   pad precision - precision ;
   92: 
   93: : zeros ( n -- )   0 max 0 ?DO  '0 emit  LOOP ;
   94: 
   95: : -zeros ( addr u -- addr' u' )
   96:   BEGIN  dup  WHILE  1- 2dup + c@ '0 <>  UNTIL  1+  THEN ;
   97: 
   98: : f$ ( f -- n )  scratch represent 0=
   99:   IF  2drop  scratch 3 min type  rdrop  EXIT  THEN
  100:   IF  '- emit  THEN ;
  101: 
  102: : f.  ( r -- ) \ float-ext f-dot
  103: \G Display (the floating-point number) @i{r} without exponent,
  104: \G followed by a space.
  105:   f$ dup >r 0<=
  106:   IF    '0 emit
  107:   ELSE  scratch r@ min type  r@ precision - zeros  THEN
  108:   '. emit r@ negate zeros
  109:   scratch r> 0 max /string 0 max -zeros type space ;
  110: \ I'm afraid this does not really implement ansi semantics wrt precision.
  111: \ Shouldn't precision indicate the number of places shown after the point?
  112: 
  113: \ Why do you think so? ANS Forth appears ambiguous on this point. -anton.
  114: 
  115: : fe. ( r -- ) \ float-ext f-e-dot
  116: \G Display @i{r} using engineering notation (with exponent dividable
  117: \G by 3), followed by a space.
  118:   f$ 1- s>d 3 fm/mod 3 * >r 1+ >r
  119:   scratch r@ tuck min tuck - >r type r> zeros
  120:   '. emit scratch r> /string type
  121:   'E emit r> . ;
  122: 
  123: : fs. ( r -- ) \ float-ext f-s-dot
  124: \G Display @i{r} using scientific notation (with exponent), followed
  125: \G by a space.
  126:   f$ 1-
  127:   scratch over c@ emit '. emit 1 /string type
  128:   'E emit . ;
  129: 
  130: : sfnumber ( c-addr u -- r true | false )
  131:     2dup [CHAR] e scan ( c-addr u c-addr2 u2 )
  132:     dup 0=
  133:     IF
  134: 	2drop 2dup [CHAR] E scan ( c-addr u c-addr3 u3 )
  135:     THEN
  136:     nip
  137:     IF
  138: 	>float
  139:     ELSE
  140: 	2drop false
  141:     THEN ;
  142: 
  143: [ifundef] compiler-notfound1
  144: defer compiler-notfound1
  145: ' no.extensions IS compiler-notfound1
  146: 
  147: :noname compiler-notfound1 execute ; is compiler-notfound
  148: 
  149: defer interpreter-notfound1
  150: ' no.extensions IS interpreter-notfound1
  151: 
  152: :noname interpreter-notfound1 execute ; is interpreter-notfound
  153: [then]
  154: 
  155: :noname ( c-addr u -- ... xt )
  156:     2dup sfnumber
  157:     IF
  158: 	2drop [comp'] FLiteral
  159:     ELSE
  160: 	defers compiler-notfound1
  161:     ENDIF ;
  162: IS compiler-notfound1
  163: 
  164: :noname ( c-addr u -- ... xt )
  165:     2dup sfnumber
  166:     IF
  167: 	2drop ['] noop
  168:     ELSE
  169: 	defers interpreter-notfound1
  170:     ENDIF ;
  171: IS interpreter-notfound1
  172: 
  173: : fvariable ( "name" -- ) \ float f-variable
  174:     Create 0.0E0 f, ;
  175:     \ does> ( -- f-addr )
  176: 
  177: 1.0e0 fasin 2.0e0 f* fconstant pi ( -- r ) \ gforth
  178: \G @code{Fconstant} -- @i{r} is the value pi; the ratio of a circle's area
  179: \G to its diameter.
  180: 
  181: : f2* ( r1 -- r2 ) \ gforth
  182:     \G Multiply @i{r1} by 2.0e0
  183:     2.0e0 f* ;
  184: 
  185: : f2/ ( r1 -- r2 ) \ gforth
  186:     \G Multiply @i{r1} by 0.5e0
  187:     0.5e0 f* ;
  188: 
  189: : 1/f ( r1 -- r2 ) \ gforth
  190:     \G Divide 1.0e0 by @i{r1}.
  191:     1.0e0 fswap f/ ;
  192: 
  193: get-current environment-wordlist set-current
  194: 1.7976931348623157e308 FConstant max-float
  195: set-current
  196: 
  197: \ We now have primitives for these, so we need not define them
  198: 
  199: \ : falog ( f -- 10^f )  [ 10.0e0 fln ] FLiteral f* fexp ;
  200: 
  201: \ : fsinh    fexpm1 fdup fdup 1.0e0 f+ f/ f+ f2/ ;
  202: \ : fcosh    fexp fdup 1/f f+ f2/ ;
  203: \ : ftanh    f2* fexpm1 fdup 2.0e0 f+ f/ ;
  204: 
  205: \ : fatanh   fdup f0< >r fabs 1.0e0 fover f- f/  f2* flnp1 f2/
  206: \            r> IF  fnegate  THEN ;
  207: \ : facosh   fdup fdup f* 1.0e0 f- fsqrt f+ fln ;
  208: \ : fasinh   fdup fdup f* 1.0e0 f+ fsqrt f/ fatanh ;
  209: 
  210: : f~abs ( r1 r2 r3 -- flag ) \ gforth
  211:     \G Approximate equality with absolute error: |r1-r2|<r3.
  212:     frot frot f- fabs fswap f< ;
  213: 
  214: : f~rel ( r1 r2 r3 -- flag ) \ gforth
  215:     \G Approximate equality with relative error: |r1-r2|<r3*|r1+r2|.
  216: 	frot frot fover fabs fover fabs f+ frot frot
  217: 	f- fabs frot frot f* f< ;
  218: 
  219: : f~ ( r1 r2 r3 -- flag ) \ float-ext f-proximate
  220:     \G ANS Forth medley for comparing r1 and r2 for equality: r3>0:
  221:     \G @code{f~abs}; r3=0: bitwise comparison; r3<0: @code{fnegate f~rel}.
  222:     fdup f0=
  223:     IF \ bitwise comparison
  224: 	fp@ float+ 1 floats over float+ over str=
  225: 	fdrop fdrop fdrop
  226: 	EXIT
  227:     THEN
  228:     fdup f0>
  229:     IF
  230: 	f~abs
  231:     ELSE
  232: 	fnegate f~rel
  233:     THEN ;
  234: 
  235: \ proposals from Krishna Myeni in <cjsp2d$47l$1@ngspool-d02.news.aol.com>
  236: \ not sure if they are a good idea
  237: 
  238: : FTRUNC ( r1 -- r2 )
  239:     \ round towards 0
  240:     \ !! should be implemented properly
  241:     F>D D>F ;
  242: 
  243: : FMOD ( r1 r2 -- r )
  244:     \ remainder of r1/r2
  245:     FOVER FOVER F/ ftrunc F* F- ;

FreeBSD-CVSweb <freebsd-cvsweb@FreeBSD.org>