|
2590
|
1 <!doctype html public "-//IETF//DTD HTML Strict//EN"> |
|
|
2 <html> |
|
|
3 <head> |
|
|
4 <title> Changes in Octave version 1 </title> |
|
|
5 </head> |
|
|
6 |
|
|
7 <body> |
|
|
8 <h1> News for Octave Version 1 </h1> |
|
|
9 <hr> |
|
|
10 |
|
|
11 <h2>Summary of changes for version 1.1.1</h2> |
|
|
12 |
|
|
13 <pre> |
|
|
14 * New built-in variables, default_return_value and |
|
|
15 define_all_return_values. |
|
|
16 |
|
|
17 If define_all_return_values is set to "false", Octave does not do |
|
|
18 anything special for return values that are left undefined, and |
|
|
19 you will get an error message if you try to use them. For |
|
|
20 example, if the function |
|
|
21 |
|
|
22 function [x, y] = f () |
|
|
23 y = 1; |
|
|
24 endfunction |
|
|
25 |
|
|
26 is called as |
|
|
27 |
|
|
28 octave:13> [a, b] = f () |
|
|
29 |
|
|
30 Octave will print an error message for the attempt to assign an |
|
|
31 undefined value to `a'. |
|
|
32 |
|
|
33 This is incompatible with Matlab, which will define the return |
|
|
34 variable `x' to be the empty matrix. To get the Matlab-like |
|
|
35 behavior, you can set the variable define_all_return_values to |
|
|
36 "true" (the default is "false") and default_return_value to `[]' |
|
|
37 (the default). Then, any return values that remain undefined when |
|
|
38 the function returns will be initialized to `[]'. |
|
|
39 |
|
|
40 If the function is called without explicitly asking for an output, |
|
|
41 it will succeed. This behavior is compatible and unchanged from |
|
|
42 previous versions of Octave. |
|
|
43 |
|
|
44 * New built-in variable suppress_verbose_help_message. If set to |
|
|
45 "true", Octave will not add additional help information to the end |
|
|
46 of the output from the help command and usage messages for |
|
|
47 built-in commands. The default value is "false". |
|
|
48 |
|
|
49 * New built-in variable PS4 is used as the prefix of echoed input |
|
|
50 (enabled with the --echo-input (-x) option). |
|
|
51 |
|
|
52 * The function size() now accepts an optional second argument. |
|
|
53 |
|
|
54 * Output from `save - ...' now goes through the pager. |
|
|
55 |
|
|
56 * The break statement may also be used to exit a function, for |
|
|
57 compatibility with Matlab. |
|
|
58 |
|
|
59 * The directory tree for installing Octave is now closer to |
|
|
60 conforming with the current GNU standards. |
|
|
61 |
|
|
62 * More bug fixes. |
|
|
63 </pre> |
|
|
64 |
|
|
65 <h2>Summary of changes for version 1.1.0</h2> |
|
|
66 |
|
|
67 <pre> |
|
|
68 * Octave now requires g++ 2.6.3 or later. This change is necessary |
|
|
69 to make template instantiations cleaner, and to avoid having to |
|
|
70 have special cases in the code for earlier versions of gcc. |
|
|
71 |
|
|
72 * A new data structure type has been added. The implementation uses |
|
|
73 an associative array with indices limited to strings, but the |
|
|
74 syntax is more like C-style structures. here are some examples of |
|
|
75 using it. |
|
|
76 |
|
|
77 Elements of structures can be of any type, including structures: |
|
|
78 |
|
|
79 octave:1> x.a = 1; |
|
|
80 octave:2> x.b = [1, 2; 3, 4]; |
|
|
81 octave:3> x.c = "string"; |
|
|
82 octave:4> x |
|
|
83 x = |
|
|
84 |
|
|
85 <structure: a b c> |
|
|
86 |
|
|
87 octave:5> x.a |
|
|
88 x.a = 1 |
|
|
89 octave:6> x.b |
|
|
90 x.b = |
|
|
91 |
|
|
92 1 2 |
|
|
93 3 4 |
|
|
94 |
|
|
95 octave:7> x.c |
|
|
96 x.c = string |
|
|
97 octave:8> x.b.d = 3 |
|
|
98 x.b.d = 3 |
|
|
99 octave:9> x.b |
|
|
100 x.b = |
|
|
101 |
|
|
102 <structure: d> |
|
|
103 |
|
|
104 octave:10> x.b.d |
|
|
105 x.b.d = 3 |
|
|
106 |
|
|
107 Functions can return structures: |
|
|
108 |
|
|
109 octave:1> a = rand (3) + rand (3) * I; |
|
|
110 octave:2> function y = f (x) |
|
|
111 > y.re = real (x); |
|
|
112 > y.im = imag (x); |
|
|
113 > endfunction |
|
|
114 octave:3> f (a) |
|
|
115 ans = |
|
|
116 |
|
|
117 <structure: im re> |
|
|
118 |
|
|
119 octave:4> ans.im |
|
|
120 ans.im = |
|
|
121 |
|
|
122 0.093411 0.229690 0.627585 |
|
|
123 0.415128 0.221706 0.850341 |
|
|
124 0.894990 0.343265 0.384018 |
|
|
125 |
|
|
126 octave:5> ans.re |
|
|
127 ans.re = |
|
|
128 |
|
|
129 0.56234 0.14797 0.26416 |
|
|
130 0.72120 0.62691 0.20910 |
|
|
131 0.89211 0.25175 0.21081 |
|
|
132 |
|
|
133 Return lists can include structure elements: |
|
|
134 |
|
|
135 octave:1> [x.u, x.s, x.v] = svd ([1, 2; 3, 4]) |
|
|
136 x.u = |
|
|
137 |
|
|
138 -0.40455 -0.91451 |
|
|
139 -0.91451 0.40455 |
|
|
140 |
|
|
141 x.s = |
|
|
142 |
|
|
143 5.46499 0.00000 |
|
|
144 0.00000 0.36597 |
|
|
145 |
|
|
146 x.v = |
|
|
147 |
|
|
148 -0.57605 0.81742 |
|
|
149 -0.81742 -0.57605 |
|
|
150 |
|
|
151 octave:8> x |
|
|
152 x = |
|
|
153 |
|
|
154 <structure: s u v> |
|
|
155 |
|
|
156 This feature should be considered experimental, but it seems to |
|
|
157 work ok. Suggestions for ways to improve it are welcome. |
|
|
158 |
|
|
159 * Octave now supports a limited form of exception handling modelled |
|
|
160 after the unwind-protect form of Lisp: |
|
|
161 |
|
|
162 unwind_protect |
|
|
163 BODY |
|
|
164 unwind_protect_cleanup |
|
|
165 CLEANUP |
|
|
166 end_unwind_protect |
|
|
167 |
|
|
168 Where BODY and CLEANUP are both optional and may contain any |
|
|
169 Octave expressions or commands. The statements in CLEANUP are |
|
|
170 guaranteed to be executed regardless of how control exits BODY. |
|
|
171 |
|
|
172 This is useful to protect temporary changes to global variables |
|
|
173 from possible errors. For example, the following code will always |
|
|
174 restore the original value of the built-in variable |
|
|
175 do_fortran_indexing even if an error occurs while performing the |
|
|
176 indexing operation. |
|
|
177 |
|
|
178 save_do_fortran_indexing = do_fortran_indexing; |
|
|
179 unwind_protect |
|
|
180 do_fortran_indexing = "true"; |
|
|
181 elt = a (idx) |
|
|
182 unwind_protect_cleanup |
|
|
183 do_fortran_indexing = save_do_fortran_indexing; |
|
|
184 end_unwind_protect |
|
|
185 |
|
|
186 Without unwind_protect, the value of do_fortran_indexing would not |
|
|
187 be restored if an error occurs while performing the indexing |
|
|
188 operation because evaluation would stop at the point of the error |
|
|
189 and the statement to restore the value would not be executed. |
|
|
190 |
|
|
191 * Recursive directory searching has been implemented using Karl |
|
|
192 Berry's kpathsea library. Directories below path elements that |
|
|
193 end in // are searched recursively for .m files. |
|
|
194 |
|
|
195 * Octave now waits for additional input when a pair of parentheses |
|
|
196 is `open' instead of giving an error. This allows one to write |
|
|
197 statements like this |
|
|
198 |
|
|
199 if (big_long_variable_name == other_long_variable_name |
|
|
200 || not_so_short_variable_name > 4 |
|
|
201 && y > x) |
|
|
202 some (code, here); |
|
|
203 |
|
|
204 without having to clutter up the if statement with continuation |
|
|
205 characters. |
|
|
206 |
|
|
207 * Continuation lines are now allowed in string constants and are |
|
|
208 handled correctly inside matrix constants. |
|
|
209 |
|
|
210 * Both `...{whitespace}\n' and `\{whitespace}\n' can be used to |
|
|
211 introduce continuation lines, where {whitespace} may include |
|
|
212 spaces, tabs and comemnts. |
|
|
213 |
|
|
214 * The script directory has been split up by topic. |
|
|
215 |
|
|
216 * Dynamic linking mostly works with dld. The following limitations |
|
|
217 are known problems: |
|
|
218 |
|
|
219 -- Clearing dynamically linked functions doesn't work. |
|
|
220 |
|
|
221 -- Dynamic linking only works with dld, which has not been ported |
|
|
222 to very many systems yet. |
|
|
223 |
|
|
224 -- Configuring with --enable-lite-kernel seems to mostly work to |
|
|
225 make nonessential built-in functions dynamically loaded, but |
|
|
226 there also seem to be some problems. For example, fsolve seems |
|
|
227 to always return info == 3. This is difficult to debug since |
|
|
228 gdb won't seem to allow breakpoints to be set inside |
|
|
229 dynamically loaded functions. |
|
|
230 |
|
|
231 -- Octave uses a lot of memory if the dynamically linked functions |
|
|
232 are compiled with -g. This appears to be a limitation with |
|
|
233 dld, and can be avoided by not using -g to compile functions |
|
|
234 that will be linked dynamically. |
|
|
235 |
|
|
236 * fft2 and ifft2 are now built-in functions. |
|
|
237 |
|
|
238 * The `&&' and `||' logical operators are now evaluated in a |
|
|
239 short-circuit fashion and work differently than the element by |
|
|
240 element operators `&' and `|'. See the Octave manual for more |
|
|
241 details. |
|
|
242 |
|
|
243 * Expressions like 1./m are now parsed as 1 ./ m, not 1. / m. |
|
|
244 |
|
|
245 * The replot command now takes the same arguments as gplot or |
|
|
246 gsplot (except ranges, which cannot be respecified with replot |
|
|
247 (yet)) so you can add additional lines to existing plots. |
|
|
248 |
|
|
249 * The hold command has been implemented. |
|
|
250 |
|
|
251 * New function `clearplot' clears the plot window. The name `clg' |
|
|
252 is aliased to `clearplot' for compatibility with Matlab. |
|
|
253 |
|
|
254 * The commands `gplot clear' and `gsplot clear' are equivalent to |
|
|
255 `clearplot'. (Previously, `gplot clear' would evaluate `clear' as |
|
|
256 an ordinary expression and clear all the visible variables.) |
|
|
257 |
|
|
258 * The Matlab-style plotting commands have been improved. They now |
|
|
259 accept line-style arguments, multiple x-y pairs, and other plot |
|
|
260 option flags. For example, |
|
|
261 |
|
|
262 plot (x, y, "@12", x, y2, x, y3, "4", x, y4, "+") |
|
|
263 |
|
|
264 results in a plot with |
|
|
265 |
|
|
266 y plotted with points of type 2 ("+") and color 1 (red). |
|
|
267 y2 plotted with lines. |
|
|
268 y3 plotted with lines of color 4. |
|
|
269 y4 plotted with points which are "+"s. |
|
|
270 |
|
|
271 the help message for `plot' and `plot_opt' provide full |
|
|
272 descriptions of the options. |
|
|
273 |
|
|
274 * NaN is now dropped from plot data, and Inf is converted to a |
|
|
275 very large value before calling gnuplot. |
|
|
276 |
|
|
277 * Improved load and save commands: |
|
|
278 |
|
|
279 -- The save and load commands can now read and write a new binary |
|
|
280 file format. Conversion to and from IEEE big and little endian |
|
|
281 formats is handled automatically. Conversion for other formats |
|
|
282 has not yet been implemented. |
|
|
283 |
|
|
284 -- The load command can now read Matlab .mat files, though it is |
|
|
285 not yet able to read sparse matrices or handle conversion for |
|
|
286 all data formats. |
|
|
287 |
|
|
288 -- The save command can write Matlab .mat files. |
|
|
289 |
|
|
290 -- The load command automatically determines the save format |
|
|
291 (binary, ascii, or Matlab binary). |
|
|
292 |
|
|
293 -- The default format for the save command is taken from the |
|
|
294 built-in variable `default_save_format'. |
|
|
295 |
|
|
296 -- The save and load commands now both accept a list of globbing |
|
|
297 patterns so you can easily load a list of variables from a |
|
|
298 file. |
|
|
299 |
|
|
300 -- The load command now accepts the option -list, for listing the |
|
|
301 variable names without actually loading the data. With |
|
|
302 -verbose, it prints a long listing. |
|
|
303 |
|
|
304 -- The load command now accepts the option -float-binary, for |
|
|
305 saving floating point data in binary files in single precision. |
|
|
306 |
|
|
307 * who and whos now accept a list of globbing patterns so you can |
|
|
308 limit the lists of variables and functions to those that match a |
|
|
309 given set of patterns. |
|
|
310 |
|
|
311 * New functions for manipulating polynomials |
|
|
312 |
|
|
313 compan -- companion matrix corresponding to polynomial coefficients |
|
|
314 conv -- convolve two vectors |
|
|
315 deconv -- deconvolve two vectors |
|
|
316 roots -- find the roots of a polynomial |
|
|
317 poly -- characteristic polynomial of a matrix |
|
|
318 polyderiv -- differentiate a polynomial |
|
|
319 polyinteg -- integrate a polynomial |
|
|
320 polyreduce -- reduce a polynomial to minimum number of terms |
|
|
321 polyval -- evaluate a polynomial at a point |
|
|
322 polyvalm -- evaluate a polynomial in the matrix sense |
|
|
323 residue -- partial fraction expansion corresponding to the ratio |
|
|
324 of two polynomials |
|
|
325 |
|
|
326 * New functions for manipulating sets |
|
|
327 |
|
|
328 create_set -- create a set of unique values |
|
|
329 complement -- find the complement of two sets |
|
|
330 intersection -- find the intersection of two sets |
|
|
331 union -- find the union of two sets |
|
|
332 |
|
|
333 * New elementary functions: |
|
|
334 |
|
|
335 acot acoth acsc acsch |
|
|
336 asec asech cot coth |
|
|
337 csc csch log2 sec |
|
|
338 sech |
|
|
339 |
|
|
340 * New special functions: |
|
|
341 |
|
|
342 beta -- beta function |
|
|
343 betai -- incomplete beta function |
|
|
344 gammai -- incomplete gamma function |
|
|
345 |
|
|
346 * New image processing functions: |
|
|
347 |
|
|
348 colormap -- set and return current colormap |
|
|
349 gray -- set a gray colormap |
|
|
350 gray2ind -- image format conversion |
|
|
351 image -- display an image |
|
|
352 imagesc -- scale and display an image |
|
|
353 imshow -- display images |
|
|
354 ind2gray -- image format conversion |
|
|
355 ind2rgb -- image format conversion |
|
|
356 loadimage -- load an image from a file |
|
|
357 ntsc2rgb -- image format conversion |
|
|
358 ocean -- set a color colormap |
|
|
359 rgb2ind -- image format conversion |
|
|
360 rgb2ntsc -- image format conversion |
|
|
361 saveimage -- save an image to a file |
|
|
362 |
|
|
363 * New time and date funcitons: |
|
|
364 |
|
|
365 tic -- set wall-clock timer |
|
|
366 toc -- get elapsed wall-clock time, since timer last set |
|
|
367 etime -- another way to get elapsed wall-clock time |
|
|
368 cputime -- get CPU time used since Octave started |
|
|
369 is_leap_year -- is the given year a leap year? |
|
|
370 |
|
|
371 * Other new functions: |
|
|
372 |
|
|
373 bug_report -- submit a bug report to the bug-octave mailing list |
|
|
374 |
|
|
375 toascii -- convert a string to a matrix of ASCII character codes |
|
|
376 |
|
|
377 octave_tmp_file -- generate a unique temporary file name |
|
|
378 |
|
|
379 undo_string_escapes -- replace special characters in a string by |
|
|
380 their backslash forms |
|
|
381 |
|
|
382 is_struct -- determine whether something is a structure data type |
|
|
383 |
|
|
384 feof -- check EOF condition for a specified file |
|
|
385 ferror -- check error state for a specified file |
|
|
386 fread -- read binary data from a file |
|
|
387 fwrite -- write binary data to a file |
|
|
388 |
|
|
389 file_in_path -- check to see if named file exists in given path |
|
|
390 |
|
|
391 kbhit -- get a single character from the terminal |
|
|
392 |
|
|
393 axis -- change plot ranges |
|
|
394 hist -- plot histograms |
|
|
395 |
|
|
396 diary -- save commands and output to a file |
|
|
397 |
|
|
398 type -- show the definition of a function |
|
|
399 which -- print the type of an identifier or the location of a |
|
|
400 function file |
|
|
401 |
|
|
402 isieee -- Returns 1 if host uses IEEE floating point |
|
|
403 realmax -- Returns largest floating point number |
|
|
404 realmin -- Returns smallest floating point number |
|
|
405 |
|
|
406 gcd -- greatest common divisor |
|
|
407 lcm -- least common multiple |
|
|
408 |
|
|
409 null -- orthonormal basis of the null space of a matrix |
|
|
410 orth -- orthonormal basis of the range space of a matrix |
|
|
411 |
|
|
412 fft2 -- two-dimensional fast fourier transform |
|
|
413 ifft2 -- two-dimensional inverse fast fourier transform |
|
|
414 filter -- digital filter |
|
|
415 fftfilt -- filter using fft |
|
|
416 fftconv -- convolve to vectors using fft |
|
|
417 sinc -- returns sin(pi*x)/(pi*x) |
|
|
418 freqz -- compute the frequency response of a filter |
|
|
419 |
|
|
420 * The meaning of nargin (== args.length ()) in built-in functions |
|
|
421 has been changed to match the meaning of nargin in user-defined |
|
|
422 functions. |
|
|
423 |
|
|
424 * Variable return lists. Octave now has a real mechanism for |
|
|
425 handling functions that return an unspecified number of values, |
|
|
426 so it is no longer necessary to place an upper bound on the number |
|
|
427 of outputs that a function can produce. |
|
|
428 |
|
|
429 Here is an example of a function that uses the new syntax to |
|
|
430 produce n values: |
|
|
431 |
|
|
432 function [...] = foo (n) |
|
|
433 for i = 1:n |
|
|
434 vr_val (i * x); |
|
|
435 endfor |
|
|
436 endfunction |
|
|
437 |
|
|
438 * New keyword, all_va_args, that allows the entire list of va_args |
|
|
439 to be passed to another function. For example, given the functions |
|
|
440 |
|
|
441 function f (...) |
|
|
442 while (nargin--) |
|
|
443 disp (va_arg ()) |
|
|
444 endwhile |
|
|
445 endfunction |
|
|
446 function g (...) |
|
|
447 f ("begin", all_va_args, "end") |
|
|
448 endfunction |
|
|
449 |
|
|
450 the statement |
|
|
451 |
|
|
452 g (1, 2, 3) |
|
|
453 |
|
|
454 prints |
|
|
455 |
|
|
456 begin |
|
|
457 1 |
|
|
458 2 |
|
|
459 3 |
|
|
460 end |
|
|
461 |
|
|
462 all_va_args may be used more than once, but can only be used |
|
|
463 within functions that take a variable number of arguments. |
|
|
464 |
|
|
465 * If given a second argument, svd now returns an economy-sized |
|
|
466 decomposition, eliminating the unecessary rows or columns of U or |
|
|
467 V. |
|
|
468 |
|
|
469 * The max and min functions correctly handle complex matrices in |
|
|
470 which some columns contain real values only. |
|
|
471 |
|
|
472 * The find function now handles 2 and 3 output arguments. |
|
|
473 |
|
|
474 * The qr function now allows computation of QR with pivoting. |
|
|
475 |
|
|
476 * hilb() is much faster for large matrices. |
|
|
477 |
|
|
478 * computer() is now a built-in function. |
|
|
479 |
|
|
480 * pinv() is now a built-in function. |
|
|
481 |
|
|
482 * The output from the history command now goes through the pager. |
|
|
483 |
|
|
484 * If a function is called without assigning the result, nargout is |
|
|
485 now correctly set to 0. |
|
|
486 |
|
|
487 * It is now possible to write functions that only set some return |
|
|
488 values. For example, calling the function |
|
|
489 |
|
|
490 function [x, y, z] = f () x = 1; z = 2; endfunction |
|
|
491 |
|
|
492 as |
|
|
493 |
|
|
494 [a, b, c] = f () |
|
|
495 |
|
|
496 produces: |
|
|
497 |
|
|
498 a = 1 |
|
|
499 |
|
|
500 b = [](0x0) |
|
|
501 |
|
|
502 c = 2 |
|
|
503 |
|
|
504 * The shell_cmd function has been renamed to system (the name |
|
|
505 shell_cmd remains for compatibility). It now returns [output, status]. |
|
|
506 |
|
|
507 * New built-in variable `OCTAVE_VERSION'. Also a new function, |
|
|
508 version, for compatibility with Matlab. |
|
|
509 |
|
|
510 * New built-in variable `automatic_replot'. If it is "true", Octave |
|
|
511 will automatically send a replot command to gnuplot each time the |
|
|
512 plot changes. Since this is fairly inefficient, the default value |
|
|
513 is "false". |
|
|
514 |
|
|
515 * New built-in variable `whitespace_in_literal_matrix' allows some |
|
|
516 control over how Octave decides to convert spaces to commas in |
|
|
517 matrix expressions like `[m (1)]'. |
|
|
518 |
|
|
519 If the value of `whitespace_in_literal_matrix' is "ignore", Octave |
|
|
520 will never insert a comma or a semicolon in a literal matrix list. |
|
|
521 For example, the expression `[1 2]' will result in an error |
|
|
522 instead of being treated the same as `[1, 2]', and the expression |
|
|
523 |
|
|
524 [ 1, 2, |
|
|
525 3, 4 ] |
|
|
526 |
|
|
527 will result in the vector [1 2 3 4] instead of a matrix. |
|
|
528 |
|
|
529 If the value of `whitespace_in_literal_matrix' is "traditional", |
|
|
530 Octave will convert spaces to a comma between identifiers and `('. |
|
|
531 For example, given the matrix |
|
|
532 |
|
|
533 m = [3 2] |
|
|
534 |
|
|
535 the expression |
|
|
536 |
|
|
537 [m (1)] |
|
|
538 |
|
|
539 will be parsed as |
|
|
540 |
|
|
541 [m, (1)] |
|
|
542 |
|
|
543 and will result in |
|
|
544 |
|
|
545 [3 2 1] |
|
|
546 |
|
|
547 and the expression |
|
|
548 |
|
|
549 [ 1, 2, |
|
|
550 3, 4 ] |
|
|
551 |
|
|
552 will result in a matrix because the newline character is converted |
|
|
553 to a semicolon (row separator) even though there is a comma at the |
|
|
554 end of the first line (trailing commas or semicolons are ignored). |
|
|
555 This is apparently how Matlab behaves. |
|
|
556 |
|
|
557 Any other value for `whitespace_in_literal_matrix' results in |
|
|
558 behavior that is the same as traditional, except that Octave does |
|
|
559 not convert spaces to a comma between identifiers and `('. |
|
|
560 For example, the expression |
|
|
561 |
|
|
562 [m (1)] |
|
|
563 |
|
|
564 will produce 3. This is the way Octave has always behaved. |
|
|
565 |
|
|
566 * Line numbers in error messages for functions defined in files and |
|
|
567 for script files now correspond to the file line number, not the |
|
|
568 number of lines after the function keyword appeared. |
|
|
569 |
|
|
570 * Octave now extracts help from script files. The comments must |
|
|
571 come before any other statements in the file. |
|
|
572 |
|
|
573 * In function files, the first block of comments in the file will |
|
|
574 now be interpreted as the help text if it doesn't look like the |
|
|
575 Octave copyright notice. Otherwise, Octave extracts the first set |
|
|
576 of comments after the function keyword. |
|
|
577 |
|
|
578 * The function clock is more accurate on systems that have the |
|
|
579 gettimeofday() function. |
|
|
580 |
|
|
581 * The standard output stream is now automatically flushed before |
|
|
582 reading from stdin with any of the *scanf() functions. |
|
|
583 |
|
|
584 * Expanded reference card. |
|
|
585 |
|
|
586 * The Octave distribution now includes a frequently asked questions |
|
|
587 file, with answers. Better answers and more questions (with |
|
|
588 answers!) are welcome. |
|
|
589 |
|
|
590 * New option --verbose. If Octave is invoked with --verbose and not |
|
|
591 --silent, a message is printed if an octaverc file is read while |
|
|
592 Octave is starting. |
|
|
593 |
|
|
594 * An improved configure script generated by Autoconf 2.0. |
|
|
595 |
|
|
596 * Lots of bug fixes. |
|
|
597 </pre> |
|
|
598 |
|
|
599 <h2>Summary of changes for version 1.0</h2> |
|
|
600 |
|
|
601 <pre> |
|
|
602 * C-style I/O functions now handle files referenced by name or by |
|
|
603 number more consistently. |
|
|
604 </pre> |
|
|
605 |
|
|
606 <h2>Summary of changes for version 0.83</h2> |
|
|
607 |
|
|
608 <pre> |
|
|
609 * Loading global symbols should work now. |
|
|
610 |
|
|
611 * Clearing the screen doesn't reprint the prompt unnecessarily. |
|
|
612 |
|
|
613 * The operations <complex scalar> OP <real matrix> for OP == +, -, |
|
|
614 *, or ./ no longer crash Octave. |
|
|
615 |
|
|
616 * More portability and configuration fixes. |
|
|
617 </pre> |
|
|
618 |
|
|
619 <h2>Summary of changes for version 0.82</h2> |
|
|
620 |
|
|
621 <pre> |
|
|
622 * Octave now comes with a reference card. |
|
|
623 |
|
|
624 * The manual has been improved, but more work remains to be done. |
|
|
625 |
|
|
626 * The atanh function now works for complex arguments. |
|
|
627 |
|
|
628 * The asin, acos, acosh, and atanh functions now work properly when |
|
|
629 given real-valued arguments that produce complex results. |
|
|
630 |
|
|
631 * SEEK_SET, SEEK_CUR, and SEEK_END are now constants. |
|
|
632 |
|
|
633 * The `using' qualifier now works with gplot and gsplot when the |
|
|
634 data to plot is coming directly from a file. |
|
|
635 |
|
|
636 * The strcmp function now works correctly for empty strings. |
|
|
637 |
|
|
638 * Eliminated bogus parse error for M-files that don't end with `end' |
|
|
639 or `endfunction'. |
|
|
640 |
|
|
641 * For empty matrices with one nonzero dimension, the +, -, .*, and |
|
|
642 ./ operators now correctly preserve the dimension. |
|
|
643 |
|
|
644 * Octave no longer crashes if you type ^D at the beginning of a line |
|
|
645 in the middle of defining a loop or if statement. |
|
|
646 |
|
|
647 * On AIX systems, Back off on indexing DiagArray via Proxy class to |
|
|
648 avoid gcc (or possibly AIX assembler?) bug. |
|
|
649 |
|
|
650 * Various other bug and portability fixes. |
|
|
651 </pre> |
|
|
652 |
|
|
653 <h2>Summary of changes for version 0.81</h2> |
|
|
654 |
|
|
655 <pre> |
|
|
656 * Octave no longer dumps core if you try to define a function in |
|
|
657 your .octaverc file. |
|
|
658 |
|
|
659 * Fixed bug in Array class that resulted in bogus off-diagonal |
|
|
660 elements when computing eigenvalue and singular value |
|
|
661 decompositions. |
|
|
662 |
|
|
663 * Fixed bug that prevented lsode from working on the SPARCstation, |
|
|
664 at least with some versions of Sun's f77. This bug was introduced |
|
|
665 in 0.80, when I changed LSODE to allow the user to abort the |
|
|
666 integration from within the RHS function. |
|
|
667 |
|
|
668 * Fixed bug that prevented global attribute of variables from being |
|
|
669 saved with save(), and another that prevented load() from working |
|
|
670 at all. |
|
|
671 </pre> |
|
|
672 |
|
|
673 <h2>Summary of changes for version 0.80</h2> |
|
|
674 |
|
|
675 <pre> |
|
|
676 * I have started working on a manual for the C++ classes. At this |
|
|
677 point, it is little more than a list of function names. If you |
|
|
678 would like to volunteer to help work on this, please contact |
|
5041
|
679 bug@octave.org. |
|
2590
|
680 |
|
|
681 * The patterns accepted by the save and clear commands now work like |
|
|
682 file name globbing patterns instead of regular expressions. I |
|
|
683 apologize for any inconvenience this change may cause, but file |
|
|
684 name globbing seems like a more reasonable style of pattern |
|
|
685 matching for this purpose. |
|
|
686 |
|
|
687 * It is now possible to specify tolerances and other optional inputs |
|
|
688 for dassl, fsolve, lsode, npsol, qpsol, and quad. For each of |
|
|
689 these functions, there is a corresponding function X_options, |
|
|
690 which takes a keyword and value arguments. If invoked without any |
|
|
691 arguments, the X_options functions print a list of possible |
|
|
692 keywords and current values. For example, |
|
|
693 |
|
|
694 npsol_options () |
|
|
695 |
|
|
696 prints a list of possible options with values, and |
|
|
697 |
|
|
698 npsol_options ("major print level", 10) |
|
|
699 |
|
|
700 sets the major print level to 10. |
|
|
701 |
|
|
702 The keyword match is not case sensitive, and the keywords may be |
|
|
703 abbreviated to the shortest unique match. For example, |
|
|
704 |
|
|
705 npsol_options ("ma p", 10) |
|
|
706 |
|
|
707 is equivalent to the statement shown above. |
|
|
708 |
|
|
709 * The new built-in variable save_precision can be used to set the |
|
|
710 number of digits preserved by the ASCII save command. |
|
|
711 |
|
|
712 * Assignment of [] now works in most cases to allow you to delete |
|
|
713 rows or columns of matrices and vectors. For example, given a |
|
|
714 4x5 matrix A, the assignment |
|
|
715 |
|
|
716 A (3, :) = [] |
|
|
717 |
|
|
718 deletes the third row of A, and the assignment |
|
|
719 |
|
|
720 A (:, 1:2:5) = [] |
|
|
721 |
|
|
722 deletes the first, third, and fifth columns. |
|
|
723 |
|
|
724 * Variable argument lists. Octave now has a real mechanism for |
|
|
725 handling functions that take an unspecified number of arguments, |
|
|
726 so it is no longer necessary to place an upper bound on the number |
|
|
727 of optional arguments that a function can accept. |
|
|
728 |
|
|
729 Here is an example of a function that uses the new syntax to print |
|
|
730 a header followed by an unspecified number of values: |
|
|
731 |
|
|
732 function foo (heading, ...) |
|
|
733 disp (heading); |
|
|
734 va_start (); |
|
|
735 while (--nargin) |
|
|
736 disp (va_arg ()); |
|
|
737 endwhile |
|
|
738 endfunction |
|
|
739 |
|
|
740 Note that the argument list must contain at least one named |
|
|
741 argument (this restriction may eventually be removed), and the |
|
|
742 ellipsis must appear as the last element of the argument list. |
|
|
743 |
|
|
744 Calling va_start() positions an internal pointer to the first |
|
|
745 unnamed argument and allows you to cycle through the arguments |
|
|
746 more than once. It is not necessary to call va_start() if you |
|
|
747 do not plan to cycle through the arguments more than once. |
|
|
748 |
|
|
749 * Recursive functions should work now. |
|
|
750 |
|
|
751 * The environment variable OCTAVE_PATH is now handled in the same |
|
|
752 way as TeX handles TEXINPUTS. If the path starts with `:', the |
|
|
753 standard path is prepended to the value obtained from the |
|
|
754 environment. If it ends with `:' the standard path is appended to |
|
|
755 the value obtained from the environment. |
|
|
756 |
|
|
757 * New functions, from Kurt Hornik (hornik@neuro.tuwien.ac.at) and |
|
|
758 the Department of Probability Theory and Statistics TU Wien, |
|
|
759 Austria: |
|
|
760 |
|
|
761 corrcoef -- corrcoef (X, Y) is the correlation between the i-th |
|
|
762 variable in X and the j-th variable in Y |
|
|
763 corrcoef (X) is corrcoef (X, X) |
|
|
764 cov -- cov (X, Y) is the covariance between the i-th |
|
|
765 variable in X and the j-th variable in Y |
|
|
766 cov (X) is cov (X, X) |
|
|
767 gls -- generalized least squares estimation |
|
|
768 kurtosis -- kurtosis(x) = N^(-1) std(x)^(-4) SUM_i (x(i)-mean(x))^4 - 3 |
|
|
769 If x is a matrix, return the row vector containing |
|
|
770 the kurtosis of each column |
|
|
771 mahalanobis -- returns Mahalanobis' D-square distance between the |
|
|
772 multivariate samples X and Y, which must have the |
|
|
773 same number of components (columns), but may have |
|
|
774 a different number of observations (rows) |
|
|
775 ols -- ordinary least squares estimation |
|
|
776 pinv -- returns the pseudoinverse of X; singular values |
|
|
777 less than tol are ignored |
|
|
778 skewness -- skewness (x) = N^(-1) std(x)^(-3) SUM_i (x(i)-mean(x))^3 |
|
|
779 if x is a matrix, return the row vector containing |
|
|
780 the skewness of each column |
|
|
781 |
|
|
782 * Errors in user-supplied functions called from dassl, fsolve, |
|
|
783 lsode, npsol, and quad are handled more gracefully. |
|
|
784 |
|
|
785 * Programming errors in the use of the C++ classes within Octave |
|
|
786 should no longer cause Octave to abort. Instead, Octave's error |
|
|
787 handler function is called and execution continues as best as is |
|
|
788 possible. This should result in eventually returning control to |
|
|
789 the top-level Octave prompt. (It would be nice to have a real |
|
|
790 exception handling mechanism...) |
|
|
791 |
|
|
792 * A number of memory leaks have been eliminated. Thanks to |
|
|
793 Fong Kin Fui <fui@ee.nus.sg> for reporting them. |
|
|
794 |
|
|
795 * The C++ matrix classes are now derived from a generic |
|
|
796 template-based array class. |
|
|
797 |
|
|
798 * The readline function operate-and-get-next (from bash) is now |
|
|
799 available and bound to C-O by default. |
|
|
800 |
|
|
801 * Octave now uses the version of readline currently distributed with |
|
|
802 bash-1.13. On some systems, interactive invocations of Octave |
|
|
803 will now blink the cursor to show matching parens. |
|
|
804 |
|
|
805 * By default, include files are now installed in |
|
|
806 $prefix/include/octave instead of $prefix/include. |
|
|
807 |
|
|
808 * Octave now uses a config.h file instead of putting all defines on |
|
|
809 the compiler command line. |
|
|
810 </pre> |
|
|
811 |
|
|
812 <h2>Summary of changes for version 0.79</h2> |
|
|
813 |
|
|
814 <pre> |
|
|
815 * New control systems functions: |
|
|
816 |
|
|
817 dgram -- Returns the discrete controllability and observability gramian. |
|
|
818 dlqr -- Discrete linear quadratic regulator design. |
|
|
819 dlqe -- Discrete linear quadratic estimator (Kalman Filter) design. |
|
|
820 c2d -- Convert continuous system description to discrete time |
|
|
821 description assuming zero-order hold and given sample time. |
|
|
822 |
|
|
823 * The max (min) functions can now return the index of the max (min) |
|
|
824 value as a second return value. |
|
|
825 </pre> |
|
|
826 |
|
|
827 <h2>Summary of changes for version 0.78</h2> |
|
|
828 |
|
|
829 <pre> |
|
|
830 * Octave's handling of global variables has been completely |
|
|
831 rewritten. To access global variables inside a function, you must |
|
|
832 now declare them to be global within the function body. Likewise, |
|
|
833 if you do not declare a variable as global at the command line, |
|
|
834 you will not have access to it within a function, even if it is |
|
|
835 declared global there. For example, given the function |
|
|
836 |
|
|
837 function f () |
|
|
838 global x = 1; |
|
|
839 y = 2; |
|
|
840 endfunction |
|
|
841 |
|
|
842 the global variable `x' is not visible at the top level until the |
|
|
843 command |
|
|
844 |
|
|
845 octave:13> global x |
|
|
846 |
|
|
847 has been evaluated, and the variable `y' remains local to the |
|
|
848 function f() even if it is declared global at the top level. |
|
|
849 |
|
|
850 Clearing a global variable at the top level will remove its global |
|
|
851 scope and leave it undefined. For example, |
|
|
852 |
|
|
853 octave:1> function f () # Define a function that accesses |
|
|
854 > global x; # the global variable `x'. |
|
|
855 > x |
|
|
856 > endfunction |
|
|
857 octave:2> global x = 1 # Give the variable `x' a value. |
|
|
858 octave:3> f () # Evaluating the function accesses the |
|
|
859 x = 1 # global `x'. |
|
|
860 octave:4> clear x # Remove `x' from global scope, clear value. |
|
|
861 octave:5> x = 2 # Define new local `x' at the top level |
|
|
862 x = 2 |
|
|
863 octave:6> f # The global `x' is no longer defined. |
|
|
864 error: `x' undefined near line 1 column 25 |
|
|
865 error: evaluating expression near line 1, column 25 |
|
|
866 error: called from `f' |
|
|
867 octave:7> x # But the local one is. |
|
|
868 x = 2 |
|
|
869 |
|
|
870 * The new function, `is_global (string)' returns 1 if the variable |
|
|
871 named by string is globally visible. Otherwise, returns 0. |
|
|
872 |
|
|
873 * The implementation of `who' has changed. It now accepts the |
|
|
874 following options: |
|
|
875 |
|
|
876 -b -builtins -- display info for built-in variables and functions |
|
|
877 -f -functions -- display info for currently compiled functions |
|
|
878 -v -variables -- display info for user variables |
|
|
879 -l -long -- display long info |
|
|
880 |
|
|
881 The long output looks like this: |
|
|
882 |
|
|
883 octave:5> who -l |
|
|
884 |
|
|
885 *** currently compiled functions: |
|
|
886 |
|
|
887 prot type rows cols name |
|
|
888 ==== ==== ==== ==== ==== |
|
|
889 wd user function - - f |
|
|
890 |
|
|
891 *** local user variables: |
|
|
892 |
|
|
893 prot type rows cols name |
|
|
894 ==== ==== ==== ==== ==== |
|
|
895 wd real scalar 1 1 y |
|
|
896 |
|
|
897 *** globally visible user variables: |
|
|
898 |
|
|
899 prot type rows cols name |
|
|
900 ==== ==== ==== ==== ==== |
|
|
901 wd complex matrix 13 13 x |
|
|
902 |
|
|
903 where the first character of the `protection' field is `w' if the |
|
|
904 symbol can be redefined, and `-' if it has read-only access. The |
|
|
905 second character may be `d' if the symbol can be deleted, or `-' |
|
|
906 if the symbol cannot be cleared. |
|
|
907 |
|
|
908 * The new built-in variable ignore_function_time_stamp can be used |
|
|
909 to prevent Octave from calling stat() each time it looks up |
|
|
910 functions defined in M-files. If set to "system", Octave will not |
|
|
911 automatically recompile M-files in subdirectories of |
|
|
912 $OCTAVE_HOME/lib/VERSION if they have changed since they were last |
|
|
913 compiled, but will recompile other M-files in the LOADPATH if they |
|
|
914 change. If set to "all", Octave will not recompile any M-files |
|
|
915 unless their definitions are removed with clear. For any other |
|
|
916 value of ignore_function_time_stamp, Octave will always check to |
|
|
917 see if functions defined in M-files need to recompiled. The |
|
|
918 default value of ignore_function_time_stamp is "system". |
|
|
919 |
|
|
920 * The new built-in variable EDITOR can be used to specify the editor |
|
|
921 for the edit_history command. It is set to the value of the |
|
|
922 environment variable EDITOR, or `vi' if EDITOR is not set, or is |
|
|
923 empty. |
|
|
924 |
|
|
925 * There is a new built-in variable, INFO_FILE, which is used as the |
|
|
926 location of the info file. Its initial value is |
|
|
927 $OCTAVE_HOME/info/octave.info, so `help -i' should now work |
|
|
928 provided that OCTAVE_HOME is set correctly, even if Octave is |
|
|
929 installed in a directory different from that specified at compile |
|
|
930 time. |
|
|
931 |
|
|
932 * There is a new command line option, --info-file FILE, that may be |
|
|
933 used to set Octave's idea of the location of the info file. It |
|
|
934 will override any value of OCTAVE_INFO_FILE found in the |
|
|
935 environment, but not any INFO_FILE="filename" commands found in |
|
|
936 the system or user startup files. |
|
|
937 |
|
|
938 * Octave's Info reader will now recognize gzipped files that have |
|
|
939 names ending in `.gz'. |
|
|
940 |
|
|
941 * The save command now accepts regular expressions as arguments. |
|
|
942 Note that these patterns are regular expressions, and do not work |
|
|
943 like filename globbing. For example, given the variables `a', |
|
|
944 `aa', and `a1', the command `save a*' saves `a' and `aa' but not |
|
|
945 `a1'. To match all variables beginning with `a', you must use an |
|
|
946 expression like `a.*' (match all sequences beginning with `a' |
|
|
947 followed by zero or more characters). |
|
|
948 |
|
|
949 * Line and column information is included in more error messages. |
|
|
950 </pre> |
|
|
951 |
|
|
952 <h2>Summary of changes for version 0.77</h2> |
|
|
953 |
|
|
954 <pre> |
|
|
955 * Improved help. The command `help -i topic' now uses the GNU Info |
|
|
956 browser to display help for the given topic directly from the |
|
|
957 Texinfo documenation. |
|
|
958 |
|
|
959 * New function: chol -- Cholesky factorization. |
|
|
960 </pre> |
|
|
961 |
|
|
962 <h2>Summary of changes for version 0.76</h2> |
|
|
963 |
|
|
964 <pre> |
|
|
965 * Better run-time error messages. Many now include line and column |
|
|
966 information indicating where the error occurred. Octave will also |
|
|
967 print a traceback for errors occurring inside functions. If you |
|
|
968 find error messages that could use improvement, or errors that |
|
|
969 Octave fails to catch, please send a bug report to |
|
5041
|
970 bug@octave.org. |
|
2590
|
971 |
|
|
972 * If gplot (or gsplot) is given a string to plot, and the string |
|
|
973 does not name a file, Octave will pass the string along to gnuplot |
|
|
974 directly. This allows commands like |
|
|
975 |
|
|
976 gplot "sin (x)" w l, data w p |
|
|
977 |
|
|
978 to work (assuming that data is a variable containing a matrix of |
|
|
979 values). |
|
|
980 |
|
|
981 * Long options (--help, --version, etc.) are supported. |
|
|
982 </pre> |
|
|
983 |
|
|
984 <h2>Summary of changes for version 0.75</h2> |
|
|
985 |
|
|
986 <pre> |
|
|
987 * The documentation is much more complete, but still could use a lot |
|
|
988 of work. |
|
|
989 |
|
|
990 * The history function now prints line numbers by default. The |
|
|
991 command `history -q' will omit them. |
|
|
992 |
|
|
993 * The clear function now accepts regular expressions. |
|
|
994 |
|
|
995 * If gplot (or gsplot) is given a string to plot, and the string |
|
|
996 names a file, Octave attempts to plot the contents of the file. |
|
|
997 |
|
|
998 * New functions: |
|
|
999 |
|
|
1000 history: |
|
|
1001 |
|
|
1002 run_history -- run commands from the history list. |
|
|
1003 edit_history -- edit commands from the history list with your |
|
|
1004 favorite editor. |
|
|
1005 |
|
|
1006 linear algebra: |
|
|
1007 |
|
|
1008 balance -- Balancing for algebraic and generalized |
|
|
1009 eigenvalue problems. |
|
|
1010 givens -- Givens rotation. |
|
|
1011 is_square -- Check to see if a matrix is square. |
|
|
1012 qzhess -- QZ decomposition of the matrix pencil (a - lambda b). |
|
|
1013 qzval -- Generalized eigenvalues for real matrices. |
|
|
1014 syl -- Sylvester equation solver. |
|
|
1015 |
|
|
1016 control systems: |
|
|
1017 |
|
|
1018 is_symmetric -- Check to see if a matrix is symmetric. |
|
|
1019 abcddim -- Check dimensions of linear dynamic system [A,B,C,D]. |
|
|
1020 is_controllable -- Check to see if [A,B,C,D] is controllable. |
|
|
1021 is_observable -- Check to see if [A,B,C,D] is observable. |
|
|
1022 are -- Solve algebraic Ricatti equation. |
|
|
1023 dare -- Solve discrete-time algebraic Ricatti equation. |
|
|
1024 lqe -- Kalman filter design for continuous linear system. |
|
|
1025 lqr -- Linear Quadratic Regulator design. |
|
|
1026 lyap -- Solve Lyapunov equation. |
|
|
1027 dlyap -- Solve discrete Lyapunov equation. |
|
|
1028 tzero -- Compute the transmission zeros of [A,B,C,D]. |
|
|
1029 </pre> |
|
|
1030 |
|
|
1031 <h2>Summary of changes for version 0.74</h2> |
|
|
1032 |
|
|
1033 <pre> |
|
|
1034 * Formal parameters to functions are now always considered to be |
|
|
1035 local variables, so things like |
|
|
1036 |
|
|
1037 global x = 0 |
|
|
1038 global y = 0 |
|
|
1039 function y = f (x) x = 1; y = x; end |
|
|
1040 f (x) |
|
|
1041 |
|
|
1042 result in the function returning 1, with the global values of x |
|
|
1043 and y unchanged. |
|
|
1044 |
|
|
1045 * Multiple assignment expressions are now allowed to take indices, |
|
|
1046 so things like |
|
|
1047 |
|
|
1048 octave:13> [a([1,2],[3,4]), b([5,6],[7,8])] = lu ([1,2;3,4]) |
|
|
1049 |
|
|
1050 will work correctly. |
|
|
1051 </pre> |
|
|
1052 |
|
|
1053 <h2>Summary of changes for version 0.73</h2> |
|
|
1054 |
|
|
1055 <pre> |
|
|
1056 * Saving and loading global variables works correctly now. |
|
|
1057 |
|
|
1058 * The save command no longer saves built-in variables. |
|
|
1059 |
|
|
1060 * Global variables are more reliable. |
|
|
1061 |
|
|
1062 * Matrices may now have one or both dimensions zero, so that |
|
|
1063 operations on empty matrices are now handled more consistently. |
|
|
1064 |
|
|
1065 By default, dimensions of the empty matrix are now printed along |
|
|
1066 with the empty matrix symbol, `[]'. For example: |
|
|
1067 |
|
|
1068 octave:13> zeros (3, 0) |
|
|
1069 ans = |
|
|
1070 |
|
|
1071 [](3x0) |
|
|
1072 |
|
|
1073 The new variable `print_empty_dimensions' controls this behavior. |
|
|
1074 |
|
|
1075 See also Carl de Boor, An Empty Exercise, SIGNUM, Volume 25, |
|
|
1076 pages 2--6, 1990, or C. N. Nett and W. M. Haddad, A |
|
|
1077 System-Theoretic Appropriate Realization of the Empty Matrix |
|
|
1078 Concept, IEEE Transactions on Automatic Control, Volume 38, |
|
|
1079 Number 5, May 1993. |
|
|
1080 |
|
|
1081 * The right and left division operators `/' and `\' will now find a |
|
|
1082 minimum norm solution if the system is not square, or if the |
|
|
1083 coefficient matrix is singular. |
|
|
1084 |
|
|
1085 * New functions: |
|
|
1086 |
|
|
1087 hess -- Hessenberg decomposition |
|
|
1088 schur -- Ordered Schur factorization |
|
|
1089 perror -- print error messages corresponding to error codes |
|
|
1090 returned from the functions fsolve, npsol, and qpsol |
|
|
1091 (with others to possibly be added later). |
|
|
1092 |
|
|
1093 * Octave now prints a warning if it finds anything other than |
|
|
1094 whitespace or comments after the final `end' or `endfunction' |
|
|
1095 statement. |
|
|
1096 |
|
|
1097 * The bodies of functions, and the for, while, and if commands are |
|
|
1098 now allowed to be empty. |
|
|
1099 |
|
|
1100 * Support for Gill and Murray's QPSOL has been added. Like NPSOL, |
|
|
1101 QPSOL is not freely redistributable either, so you must obtain |
|
|
1102 your own copy to be able to use this feature. More information |
|
|
1103 about where to find QPSOL and NPSOL are in the file README.NLP. |
|
|
1104 </pre> |
|
|
1105 |
|
|
1106 <h2>Summary of changes for version 0.72</h2> |
|
|
1107 |
|
|
1108 <pre> |
|
|
1109 * For numeric output, columns are now lined up on the decimal point. |
|
|
1110 (This requires libg++-2.3.1 or later to work correctly). |
|
|
1111 |
|
|
1112 * If octave is running interactively and the output intended for the |
|
|
1113 screen is longer than one page and a pager is available, it is |
|
|
1114 sent to the pager through a pipe. You may specify the program to |
|
|
1115 use as the pager by setting the variable PAGER. PAGER may also |
|
|
1116 specify a command pipeline. |
|
|
1117 |
|
|
1118 * Spaces are not always significant inside square brackets now, so |
|
|
1119 commands like |
|
|
1120 |
|
|
1121 [ linspace (1, 2) ] |
|
|
1122 |
|
|
1123 will work. However, some possible sources of confusion remain |
|
|
1124 because Octave tries (possibly too hard) to determine exactly what |
|
|
1125 operation is intended from the context surrounding an operator. |
|
|
1126 For example: |
|
|
1127 |
|
|
1128 -- In the command |
|
|
1129 |
|
|
1130 [ 1 - 1 ] |
|
|
1131 |
|
|
1132 the `-' is treated as a binary operator and the result is the |
|
|
1133 scalar 0, but in the command |
|
|
1134 |
|
|
1135 [ 1 -1 ] |
|
|
1136 |
|
|
1137 the `-' is treated as a unary operator and the result is the |
|
|
1138 vector [ 1 -1 ]. |
|
|
1139 |
|
|
1140 -- In the command |
|
|
1141 |
|
|
1142 a = 1; [ 1 a' ] |
|
|
1143 |
|
|
1144 the single quote character `'' is treated as a transpose operator |
|
|
1145 and the result is the vector [ 1 1 ], but in the command |
|
|
1146 |
|
|
1147 a = 1; [ 1 a ' ] |
|
|
1148 |
|
|
1149 an error message indicating an unterminated string constant is |
|
|
1150 printed. |
|
|
1151 |
|
|
1152 * Assignments are just expressions now, so they are valid anywhere |
|
|
1153 other expressions are. This means that things like |
|
|
1154 |
|
|
1155 if (a = n < m) ... endif |
|
|
1156 |
|
|
1157 are valid. This is parsed as: compare `n < m', assign the result |
|
|
1158 to the variable `a', and use it as the test expression in the if |
|
|
1159 statement. |
|
|
1160 |
|
|
1161 To help avoid errors where `=' has been used but `==' was |
|
|
1162 intended, Octave issues a warning suggesting parenthesis around |
|
|
1163 assignments used as truth values. You can suppress this warning |
|
|
1164 by adding parenthesis, or by setting the value of the new built-in |
|
|
1165 variable `warn_assign_as_truth_value' to 'false' (the default |
|
|
1166 value is 'true'). |
|
|
1167 |
|
|
1168 This is also true for multiple assignments, so expressions like |
|
|
1169 |
|
|
1170 [a, b, c] = [u, s, v] = expression |
|
|
1171 |
|
|
1172 are now possible. If the expression is a function, nargout is set |
|
|
1173 to the number of arguments for the right-most assignment. The |
|
|
1174 other assignments need not contain the same number of elements. |
|
|
1175 Extra left hand side variables in an assignment become undefined. |
|
|
1176 |
|
|
1177 * The default line style for plots is now `lines' instead of |
|
|
1178 `points'. To change it, use the `set data style STYLE' command. |
|
|
1179 |
|
|
1180 * New file handling and I/O functions: |
|
|
1181 |
|
|
1182 fopen -- open a file for reading or writing |
|
|
1183 fclose -- close a file |
|
|
1184 fflush -- flush output to a file |
|
|
1185 fgets -- read characters from a file |
|
|
1186 frewind -- set file position to the beginning of a file |
|
|
1187 fseek -- set file position |
|
|
1188 ftell -- tell file position |
|
|
1189 freport -- print a report for all open files |
|
|
1190 fscanf -- read from a file |
|
|
1191 sscanf -- read from a string |
|
|
1192 scanf -- read from the standard input |
|
|
1193 |
|
|
1194 * New built-in variables for file and I/O functions: |
|
|
1195 |
|
|
1196 stdin -- file number corresponding to the standard input stream. |
|
|
1197 stdout -- file number corresponding to the standard output stream. |
|
|
1198 stderr -- file number corresponding to the standard error stream. |
|
|
1199 |
|
|
1200 The following may be used as the final (optional) argument for |
|
|
1201 fseek: |
|
|
1202 |
|
|
1203 SEEK_SET -- set position relative to the beginning of the file. |
|
|
1204 SEEK_CUR -- set position relative to the current position. |
|
|
1205 SEEK_END -- set position relative to the end of the file. |
|
|
1206 |
|
|
1207 * New function: setstr -- convert vectors or scalars to strings |
|
|
1208 (doesn't work for matrices yet). |
|
|
1209 |
|
|
1210 * If possible, computer now prints the system type instead of |
|
|
1211 always printing `Hi Dave, I'm a HAL-9000'. |
|
|
1212 |
|
|
1213 * Octave now properly saves and restores its internal state |
|
|
1214 correctly in more places. Interrupting Octave while it is |
|
|
1215 executing a script file no longer causes it to exit. |
|
|
1216 |
|
|
1217 * Octave now does tilde expansion on each element of the LOADPATH. |
|
|
1218 |
|
|
1219 * A number of memory leaks have been plugged. |
|
|
1220 |
|
|
1221 * Dependencies for C++ source files are now generated automatically |
|
|
1222 by g++. |
|
|
1223 |
|
|
1224 * There is a new command line option, -p PATH, that may be used to |
|
|
1225 set Octave's loadpath from the command line. It will override any |
|
|
1226 value of OCTAVE_PATH found in the environment, but not any |
|
|
1227 LOADPATH="path" commands found in the system or user startup files. |
|
|
1228 |
|
|
1229 * It is now possible to override Octave's default idea of the |
|
|
1230 location of the system-wide startup file (usually stored in |
|
|
1231 $(prefix)/lib/octave/octaverc) using the environment variable |
|
|
1232 OCTAVE_HOME. If OCTAVE_HOME has a value, Octave will look for |
|
|
1233 octaverc and its M-files in the directory $OCTAVE_HOME/lib/octave. |
|
|
1234 |
|
|
1235 This allows people who are using binary distributions (as is |
|
|
1236 common with systems like Linux) to install the real octave binary |
|
|
1237 in any directory (using a name like octave.bin) and then install |
|
|
1238 a simple script like this |
|
|
1239 |
|
|
1240 #!/bin/sh |
|
|
1241 OCTAVE_HOME=/foo/bar/baz |
|
|
1242 export OCTAVE_HOME |
|
|
1243 exec octave.bin |
|
|
1244 |
|
|
1245 to be invoked as octave. |
|
|
1246 |
|
|
1247 </pre> |
|
|
1248 |
|
|
1249 <h2>Summary of changes for version 0.71</h2> |
|
|
1250 |
|
|
1251 <pre> |
|
|
1252 * Much improved plotting facility. With this release, Octave does |
|
|
1253 not require a specially modified version of gnuplot, so gnuplot |
|
|
1254 sources are no longer distributed with Octave. For a more |
|
|
1255 detailed description of the new plotting features, see the file |
|
|
1256 PLOTTING. |
|
|
1257 |
|
|
1258 * New plotting commands: |
|
|
1259 |
|
|
1260 plot -- 2D plots |
|
|
1261 semilogx -- 2D semilog plot with logscale on the x axis |
|
|
1262 semilogy -- 2D semilog plot with logscale on the y axis |
|
|
1263 loglog -- 2D log-log plot |
|
|
1264 mesh -- 3D mesh plot |
|
|
1265 meshdom -- create matrices for 3D plotting from two vectors |
|
|
1266 contour -- contour plots of 3D data |
|
|
1267 bar -- create bar graphs |
|
|
1268 stairs -- create stairstep plots |
|
|
1269 polar -- 2D plots from theta-R data |
|
|
1270 grid -- turn plot grid lines on or off |
|
|
1271 xlabel, ylabel -- place labels on the x and y axes of 2D plots |
|
|
1272 sombrero -- demonstrate 3D plotting |
|
|
1273 gplot -- 2D plot command with gnuplot-like syntax |
|
|
1274 gsplot -- 3D plot command with gnuplot-like syntax |
|
|
1275 set -- set plot options with gnuplot syntax |
|
|
1276 show -- show plot options with gnuplot syntax |
|
|
1277 closeplot -- close stream to gnuplot process |
|
|
1278 purge_tmp_files -- delete temporary files created by plot command |
|
|
1279 |
|
|
1280 * Other new commands: |
|
|
1281 |
|
|
1282 ls, dir -- print a directory listing |
|
|
1283 shell_cmd -- execute shell commands |
|
|
1284 keyboard -- get input from keyboard, useful for debugging |
|
|
1285 menu -- display a menu of options and ask for input |
|
|
1286 fft -- fast fourier transform |
|
|
1287 ifft -- inverse fast fourier transform |
|
|
1288 |
|
|
1289 * Strings may be enclosed in either single or double quote |
|
|
1290 characters. Double quote characters are not special within single |
|
|
1291 quote strings, and single quotes are not special within double |
|
|
1292 quote strings. |
|
|
1293 |
|
|
1294 * Command name completion now works for M-file names too. |
|
|
1295 |
|
|
1296 * Better help and usage messages for many functions. |
|
|
1297 |
|
|
1298 * Help is now available for functions defined in M-files. The first |
|
|
1299 block of comments is taken as the text of the help message. |
|
|
1300 |
|
|
1301 * Numerous changes in preparation to support dynamic loading of |
|
|
1302 object files with dld. |
|
|
1303 |
|
|
1304 * Bug fixes to make solving DAEs with dassl actually work. |
|
|
1305 |
|
|
1306 * The command `save file' now saves all variables in the named file. |
|
|
1307 |
|
|
1308 * If do_fortran_indexing is 'true', indexing a scalar with |
|
|
1309 [1,1,1,...] (n times) replicates its value n times. The |
|
|
1310 orientation of the resulting vector depends on the value of |
|
|
1311 prefer_column_vectors. |
|
|
1312 |
|
|
1313 * Things like [[1,2][3,4]] no longer cause core dumps, and invalid |
|
|
1314 input like [1,2;3,4,[5,6]] now produces a diagnositic message. |
|
|
1315 |
|
|
1316 * The cd, save, and load commands now do tilde expansion. |
|
|
1317 |
|
|
1318 * It's now possible to clear global variables and functions by name. |
|
|
1319 |
|
|
1320 * Use of clear inside functions is now a parse error. |
|
|
1321 </pre> |
|
|
1322 |
|
|
1323 <h2>Summary of changes for version 0.70</h2> |
|
|
1324 |
|
|
1325 <pre> |
|
|
1326 * Better parse error diagnostics. For interactive input, you get |
|
|
1327 messages like |
|
|
1328 |
|
|
1329 octave:1> a = 3 + * 4; |
|
|
1330 |
|
|
1331 parse error: |
|
|
1332 |
|
|
1333 a = 3 + * 4; |
|
|
1334 ^ |
|
|
1335 |
|
|
1336 and for script files, the message includes the file name and input |
|
|
1337 line number: |
|
|
1338 |
|
|
1339 octave:1> foo |
|
|
1340 |
|
|
1341 parse error near line 4 of file foo.m: |
|
|
1342 |
|
|
1343 a = 3 + * 4; |
|
|
1344 ^ |
|
|
1345 |
|
|
1346 * New built-in variable PS2 which is used as the secondary prompt. |
|
|
1347 The default value is '> '. |
|
|
1348 |
|
|
1349 * New file, octave-mode.el, for editing Octave code with GNU Emacs. |
|
|
1350 This is a modified version of Matthew R. Wette's matlab-mode.el. |
|
|
1351 |
|
|
1352 * Better support for missing math functions. |
|
|
1353 |
|
|
1354 * User preferences are now cached in a global struct so we don't |
|
|
1355 have to do a symbol table lookup each time we need to know what |
|
|
1356 they are. This should mean slightly improved performance for |
|
|
1357 evaluating expressions. |
|
|
1358 </pre> |
|
|
1359 |
|
|
1360 <h2>Summary of changes for version 0.69</h2> |
|
|
1361 |
|
|
1362 <pre> |
|
|
1363 * Multiple assignments are now possible, so statements like |
|
|
1364 |
|
|
1365 a = b = c = 3; |
|
|
1366 a = b = c = [1,2;3,4]; |
|
|
1367 |
|
|
1368 or |
|
|
1369 |
|
|
1370 c = (a = (b = 2) * 3 + 4) * 5 |
|
|
1371 |
|
|
1372 are legal, as are things that have even more bizarre effects, like |
|
|
1373 |
|
|
1374 a(4:6,4:6) = b(2:3,2:3) = [1,2;3,4]; |
|
|
1375 |
|
|
1376 (try it). |
|
|
1377 |
|
|
1378 * Improved parsing of strings (but they still don't work as matrix |
|
|
1379 elements). |
|
|
1380 |
|
|
1381 * An M-file may now either define a function or be a list of |
|
|
1382 commands to execute. |
|
|
1383 |
|
|
1384 * Better detection and conditional compilation of IEEE functions |
|
|
1385 isinf, finite, and isnan. |
|
|
1386 |
|
|
1387 * Replacements for acosh, asinh, atanh, and gamma from the BSD math |
|
|
1388 library for those systems that don't have them. |
|
|
1389 </pre> |
|
|
1390 |
|
|
1391 <h2>Summary of changes for version 0.68</h2> |
|
|
1392 |
|
|
1393 <pre> |
|
|
1394 * New functions: |
|
|
1395 |
|
|
1396 eval -- evaluate a string as a sequence of Octave commands. |
|
|
1397 input -- print a prompt and get user input. |
|
|
1398 </pre> |
|
|
1399 |
|
|
1400 <h2>Summary of changes for version 0.67</h2> |
|
|
1401 |
|
|
1402 <pre> |
|
|
1403 * New functions: |
|
|
1404 |
|
|
1405 find -- return the indices of nonzero elements. |
|
|
1406 |
|
|
1407 * Zero-one style indexing now works. For example, |
|
|
1408 |
|
|
1409 a = [1,2,3,4]; |
|
|
1410 b = a([1,0,0,1]) |
|
|
1411 |
|
|
1412 sets b to the first and fourth elememnts of a. |
|
|
1413 |
|
|
1414 Zero-one style indexing also works for indexing the left hand side |
|
|
1415 of an assignment. For example, |
|
|
1416 |
|
|
1417 a = rand (1,2;3,4); |
|
|
1418 a([0,1],:) = [-1,-2] |
|
|
1419 |
|
|
1420 sets the second row of a to [-1 -2] |
|
|
1421 |
|
|
1422 The behavior for the ambiguous case |
|
|
1423 |
|
|
1424 a = [1,2,3,4]; |
|
|
1425 b = a([1,1,1,1]); |
|
|
1426 |
|
|
1427 is controlled by the new global variable `prefer_zero_one_indexing'. |
|
|
1428 If this variable is equal to 'true', b will be set to [1 2 3 4]. |
|
|
1429 If it is false, b will be set to [1 1 1 1]. The default value is |
|
|
1430 'false'. |
|
|
1431 |
|
|
1432 * Using the new global variable `propagate_empty_matrices', it is |
|
|
1433 possible to have unary andy binary operations on empty matrices |
|
|
1434 return an empty matrix. The default value of this variable is |
|
|
1435 'warn', so that empty matrices are propagated but you get a |
|
|
1436 warning. Some functions, like eig and svd have also been changed |
|
|
1437 to handle this. |
|
|
1438 |
|
|
1439 * Empty matrices can be used in conditionals, but they always |
|
|
1440 evaluate to `false'. With propagate_empty_matrices = 'true', both |
|
|
1441 of the following expressions print 0: |
|
|
1442 |
|
|
1443 if [], 1, else 0, end |
|
|
1444 if ~[], 1, else 0, end |
|
|
1445 |
|
|
1446 * Octave no longer converts input like `3.2 i' or `3 I' to complex |
|
|
1447 constants directly because that causes problems inside square |
|
|
1448 brackets, where spaces are important. This abbreviated notation |
|
|
1449 *does* work if there isn't a space between the number and the i, |
|
|
1450 I, j, or J. |
|
|
1451 </pre> |
|
|
1452 |
|
|
1453 <h2>Summary of changes for version 0.66</h2> |
|
|
1454 |
|
|
1455 <pre> |
|
|
1456 * Logical unary not operator (~ or !) now works for complex. |
|
|
1457 |
|
|
1458 * Left division works. |
|
|
1459 |
|
|
1460 * Right and left element by element division should work correctly |
|
|
1461 now. |
|
|
1462 |
|
|
1463 * Numbers like .3e+2 are no longer errors. |
|
|
1464 |
|
|
1465 * Indexing a matrix with a complex value doesn't cause a core dump. |
|
|
1466 |
|
|
1467 * The min and max functions should work correctly for two arguments. |
|
|
1468 |
|
|
1469 * Improved (I hope!) configuration checks. |
|
|
1470 |
|
|
1471 * Octave is now installed as octave-M.N, where M and N are version |
|
|
1472 numbers, and octave is a link to that file. This makes it |
|
|
1473 possible to have more than one version of the interpreter installed. |
|
|
1474 </pre> |
|
|
1475 |
|
|
1476 <h2>Summary of changes for version 0.63</h2> |
|
|
1477 |
|
|
1478 <pre> |
|
|
1479 * The reshape function works again. |
|
|
1480 |
|
|
1481 * Octave now converts input like `3.2i' or `3 I' or `2.3e5 j' to be |
|
|
1482 complex constants directly, rather than requiring an expression |
|
|
1483 like `3.3 * i' to be evaluated. |
|
|
1484 </pre> |
|
|
1485 |
|
|
1486 <h2>Summary of changes for version 0.61</h2> |
|
|
1487 |
|
|
1488 <pre> |
|
|
1489 * Octave has been successfully compiled using gcc 2.3.3 and libg++ 2.3. |
|
|
1490 on a 486 system running Linux. |
|
|
1491 |
|
|
1492 * The win_texas_lotto function is now called texas_lotto (it's a |
|
|
1493 script file, and win_texas_lotto.m is too long for some Linux and |
|
|
1494 System V systems). |
|
|
1495 </pre> |
|
|
1496 |
|
|
1497 <h2>Summary of changes for version 0.57</h2> |
|
|
1498 |
|
|
1499 <pre> |
|
|
1500 * The C-like formatted print functions printf, fprintf, and sprintf |
|
|
1501 finally work. |
|
|
1502 </pre> |
|
|
1503 |
|
|
1504 <h2>Summary of changes for version 0.56</h2> |
|
|
1505 |
|
|
1506 <pre> |
|
|
1507 * By default, octave prints a short disclaimer when it starts. |
|
|
1508 (You can suppress it by invoking octave with -q). |
|
|
1509 |
|
|
1510 * You can keep octave from reading your ~/.octaverc and .octaverc |
|
|
1511 files by invoking it with -f. |
|
|
1512 |
|
|
1513 * When returning two values, eig now returns [v, d] instead of |
|
|
1514 [lambda, v], where d is a diagonal matrix made from lambda. |
|
|
1515 |
|
|
1516 * The win_texas_lotto function now produces a sorted list. |
|
|
1517 |
|
|
1518 * New functions: |
|
|
1519 |
|
|
1520 expm -- matrix exponential. |
|
|
1521 logm -- matrix logarithm. |
|
|
1522 </pre> |
|
|
1523 |
|
|
1524 <h2>Summary of changes for version 0.55</h2> |
|
|
1525 |
|
|
1526 <pre> |
|
|
1527 * The following (C-style) backslash escape sequences work in quoted |
|
|
1528 strings (useful(?) with printf()): |
|
|
1529 |
|
|
1530 \a bell \r carriage return |
|
|
1531 \b backspace \t horizontal tab |
|
|
1532 \f formfeed \v vertical tab |
|
|
1533 \n newline \\ backslash |
|
|
1534 |
|
|
1535 * Use of `...' at the end of a line will allow a statement to |
|
|
1536 continue over more than one line. |
|
|
1537 |
|
|
1538 * The names `inf' and `nan' are now aliases for `Inf' and `NaN', |
|
|
1539 respectively. |
|
|
1540 |
|
|
1541 * New functions: |
|
|
1542 |
|
|
1543 casesen -- print a warning if the luser tries to turn off case |
|
|
1544 sensitivity. |
|
|
1545 median -- find median value. |
|
|
1546 norm -- compute the norm of a matrix. |
|
|
1547 sort -- sort columns. |
|
|
1548 |
|
|
1549 * New variable, `silent_functions'. If silent_functions == 'true', |
|
|
1550 the results of expressions are not printed even if they are not |
|
|
1551 followed by a semicolon. The disp() and printf() functions still |
|
|
1552 result in output. The default value for this variable is 'false'. |
|
|
1553 |
|
|
1554 * New variable `return_last_value_computed'. If it is 'true', |
|
|
1555 functions defined in script files return the last value computed |
|
|
1556 if a return value has not been explicitly declared. The default |
|
|
1557 value for this variable is 'false'. |
|
|
1558 </pre> |
|
|
1559 |
|
|
1560 <h2>Summary of changes for version 0.52</h2> |
|
|
1561 |
|
|
1562 <pre> |
|
|
1563 * Name completion works for function and variable names currently in |
|
|
1564 the symbol tables. Coming soon: completion for names of functions |
|
|
1565 defined in script files but not yet compiled. |
|
|
1566 |
|
|
1567 * The initial value of do_fortran_indexing is now false, and the |
|
|
1568 initial value of prefer_column_vectors is now true. Swap the |
|
|
1569 values of these variables if you want behavior that is more like |
|
|
1570 Matlab. |
|
|
1571 |
|
|
1572 * All script files check the number of input arguments before doing |
|
|
1573 much real work. |
|
|
1574 |
|
|
1575 * The identifiers `i' and `j' are now also names for sqrt(-1). |
|
|
1576 These symbols may be used for other purposes, but their original |
|
|
1577 definition will reappear if they are cleared. |
|
|
1578 |
|
|
1579 * The symbol tables are now implemented with hash tables for faster |
|
|
1580 searching. |
|
|
1581 |
|
|
1582 * A small amount of help is now available for most built-in |
|
|
1583 operators, keywords and functions. Coming soon: help for script |
|
|
1584 files. |
|
|
1585 |
|
|
1586 * Without any arguments, the help command now lists all known |
|
|
1587 built-in operators, keywords and functions. |
|
|
1588 |
|
|
1589 * Generic parse errors are now signalled by `Eh, what's up doc?', |
|
|
1590 which is closer to what Bugs actually says. |
|
|
1591 |
|
|
1592 * The who command now only prints variable names by default. |
|
|
1593 Use the -fcn (or -fcns, or -functions) switch to print the names of |
|
|
1594 built-in or currently compiled functions. |
|
|
1595 </pre> |
|
|
1596 |
|
|
1597 <h2>Summary of changes for version 0.51</h2> |
|
|
1598 |
|
|
1599 <pre> |
|
|
1600 * Major overhaul of array indexing. |
|
|
1601 |
|
|
1602 * The colloc function actually works now. |
|
|
1603 </pre> |
|
|
1604 |
|
|
1605 <h2>Summary of changes for version 0.50</h2> |
|
|
1606 |
|
|
1607 <pre> |
|
|
1608 * The lsode and dassl functions now return the states only, |
|
|
1609 instead of the time and the states, so you must keep track of |
|
|
1610 the corresponding times (this is easy though, because you have |
|
|
1611 to specify a vector of desired output times anyway). |
|
|
1612 |
|
|
1613 * Solution of NLPs with NPSOL now works on the SPARC. |
|
|
1614 |
|
|
1615 * New keywords `endif', `endfor', `endfunction', `endif', and |
|
|
1616 `endwhile', which allow for better diagnostics. The `end' keyword |
|
|
1617 is still recognized. All script files have been changed to use |
|
|
1618 these new keywords in place of `end'. |
|
|
1619 |
|
|
1620 * It is now possible to uninstall Octave by doing a `make uninstall' |
|
|
1621 in the top level directory. |
|
|
1622 |
|
|
1623 * The Makefiles are much closer to conforming with GNU coding standards. |
|
|
1624 |
|
|
1625 * New functions: |
|
|
1626 |
|
|
1627 win_texas_lotto -- produce six unique random numbers between 1 and 50. |
|
|
1628 quad -- numerical integration. |
|
|
1629 lu -- LU factorization |
|
|
1630 qr -- QR factorization |
|
|
1631 dassl -- Solution of DAEs using DASSL. |
|
|
1632 |
|
|
1633 * New files: |
|
|
1634 |
|
|
1635 THANKS -- A list of people and organazations who have supported |
|
|
1636 the development of Octave. |
|
|
1637 |
|
|
1638 NEWS -- This file, listing recent changes. |
|
|
1639 |
|
|
1640 * Help is now available at the gnuplot prompt. |
|
|
1641 </pre> |
|
|
1642 |
|
|
1643 </body> |
|
|
1644 </html> |
