Mercurial > octave-nkf
annotate PROJECTS @ 12312:b10ea6efdc58 release-3-4-x ss-3-3-91
version is now 3.3.91
| author | John W. Eaton <jwe@octave.org> |
|---|---|
| date | Mon, 31 Jan 2011 08:36:58 -0500 |
| parents | b4d2080b6df7 |
| children | 83a37dc986f8 |
| rev | line source |
|---|---|
| 3162 | 1 <html> |
| 2 <pre> | |
| 2330 | 3 Octave PROJECTS -*- text -*- |
| 4 =============== | |
| 5 | |
| 8921 | 6 Look in the Octave source archive on Savannah for a possibly more |
| 7 current copy. Also, if you start working steadily on a project, | |
| 8 please let maintainers@octave.org know. We might have information | |
| 9 that could help you. You should also read the Contributing Guidelines | |
| 10 chapter in the Octave manual. | |
| 2330 | 11 |
| 12 This list is not exclusive -- there are many other things that might | |
| 8921 | 13 be good projects, but it might instead be something we already have. |
| 14 Also, some of the following items may not actually be considered good | |
| 15 ideas now. So please check with maintainers@octave.org before you | |
| 16 start working on some large project. | |
| 17 | |
| 2330 | 18 |
| 19 --------- | |
| 20 Numerical: | |
| 21 --------- | |
| 22 | |
| 23 * Improve logm, and sqrtm. | |
| 24 | |
| 3713 | 25 * Improve complex mapper functions. See W. Kahan, ``Branch Cuts for |
| 26 Complex Elementary Functions, or Much Ado About Nothing's Sign | |
| 27 Bit'' (in The State of the Art in Numerical Analysis, eds. Iserles | |
| 28 and Powell, Clarendon Press, Oxford, 1987) for explicit | |
| 29 trigonometric formulae. | |
| 2330 | 30 |
| 31 * Make functions like gamma() return the right IEEE Inf or NaN | |
| 32 values for extreme args or other undefined cases. | |
| 33 | |
| 6628 | 34 * Improve sqp. |
| 2330 | 35 |
| 36 * Fix CollocWt to handle Laguerre polynomials. Make it easy to | |
| 37 extend it to other polynomial types. | |
| 38 | |
| 39 * Add optional arguments to colloc so that it's not restricted to | |
| 40 Legendre polynomials. | |
| 41 | |
| 42 * Fix eig to also be able to solve the generalized eigenvalue | |
| 43 problem, and to solve for eigenvalues and eigenvectors without | |
| 44 performing a balancing step first. | |
| 45 | |
| 46 * Move rand, eye, xpow, xdiv, etc., functions to the matrix classes. | |
| 47 | |
| 2477 | 48 * Use octave_allocator for memory management in Array classes once |
| 49 g++ supports static member templates. | |
| 50 | |
| 2330 | 51 * Improve design of ODE, DAE, classes. |
| 52 | |
| 3141 | 53 * Make QR more memory efficient for large matrices when not all the |
| 54 columns of Q are required (apparently this is not handled by the | |
| 55 lapack code yet). | |
| 56 | |
| 5164 | 57 --------------- |
| 58 Sparse Matrices: | |
| 59 --------------- | |
| 60 | |
| 5610 | 61 * Improve QR factorization functions, using idea based on CSPARSE |
| 62 cs_dmsol.m | |
| 5164 | 63 |
|
10527
b4d2080b6df7
Replace nzmax by nnz as needed
David Bateman <dbateman@free.fr>
parents:
9213
diff
changeset
|
64 * Improve QR fqctorization by replace CXSPARSE code with SPQR code, and make |
|
b4d2080b6df7
Replace nzmax by nnz as needed
David Bateman <dbateman@free.fr>
parents:
9213
diff
changeset
|
65 the linear solve return 2-norm solutions for ill-conditioned matrices |
|
b4d2080b6df7
Replace nzmax by nnz as needed
David Bateman <dbateman@free.fr>
parents:
9213
diff
changeset
|
66 based on this new code |
|
b4d2080b6df7
Replace nzmax by nnz as needed
David Bateman <dbateman@free.fr>
parents:
9213
diff
changeset
|
67 |
| 5610 | 68 * Implement fourth argument to the sprand and sprandn, and addition |
| 69 arguments to sprandsym that the leading brand implements. | |
| 5164 | 70 |
| 5282 | 71 * Sparse logical indexing in idx_vector class so that something like |
|
10527
b4d2080b6df7
Replace nzmax by nnz as needed
David Bateman <dbateman@free.fr>
parents:
9213
diff
changeset
|
72 'a=sprandn(1e6,1e6,1e-6); a(a<1) = 0' won't cause a memory overflow. |
| 5164 | 73 |
| 74 * Other missing Functions | |
| 75 - symmmd Superseded by symamd | |
| 76 - colmmd Superseded by colamd | |
| 77 - cholinc | |
| 6627 | 78 - bicg Can this be taken from octave-forge? |
| 5164 | 79 - gmres |
| 80 - lsqr | |
| 81 - minres | |
| 82 - qmr | |
| 83 - symmlq | |
| 84 | |
| 2330 | 85 ------- |
| 86 Strings: | |
| 87 ------- | |
| 88 | |
| 2789 | 89 * Improve performance of string functions, particularly for |
| 90 searching and replacing. | |
| 91 | |
| 2330 | 92 * Make find work for strings. |
| 93 | |
| 2378 | 94 * Consider making octave_print_internal() print some sort of text |
| 95 representation for unprintable characters instead of sending them | |
| 96 directly to the terminal. (But don't do this for fprintf!) | |
| 97 | |
| 98 * Consider changing the default value of `string_fill_char' from SPC | |
| 99 to NUL. | |
| 100 | |
| 2330 | 101 ---------------- |
| 102 Other Data Types: | |
| 103 ---------------- | |
| 104 | |
| 105 * Template functions for mixed-type ops. | |
| 106 | |
|
7789
82be108cc558
First attempt at single precision tyeps
David Bateman <dbateman@free.fr>
parents:
7787
diff
changeset
|
107 * Convert other functions for use with the floating point type |
| 8921 | 108 including quad, dasrt, daspk, etc. |
|
7789
82be108cc558
First attempt at single precision tyeps
David Bateman <dbateman@free.fr>
parents:
7787
diff
changeset
|
109 |
| 2330 | 110 ------------ |
| 111 Input/Output: | |
| 112 ------------ | |
| 113 | |
| 114 * Make fread and fwrite work for complex data. Iostreams based | |
| 115 versions of these functions would also be nice, and if you are | |
| 116 working on them, it would be good to support other size | |
| 117 specifications (integer*2, etc.). | |
| 118 | |
| 119 * Move some pr-output stuff to liboctave. | |
| 120 | |
| 121 * Make the cutoff point for changing to packed storage a | |
| 122 user-preference variable with default value 8192. | |
| 123 | |
| 124 * Complain if there is not enough disk space available (I think | |
| 125 there is simply not enough error checking in the code that handles | |
| 126 writing data). | |
| 127 | |
| 128 * Make it possible to tie arbitrary input and output streams | |
| 129 together, similar to the way iostreams can be tied together. | |
| 130 | |
| 131 ----------- | |
| 132 Interpreter: | |
| 133 ----------- | |
| 134 | |
| 3162 | 135 * Allow customization of the debug prompt. |
| 136 | |
| 2601 | 137 * Fix the parser so that |
| 138 | |
| 3081 | 139 if (expr) 'this is a string' end |
| 140 | |
| 141 is parsed as IF expr STRING END. | |
| 142 | |
| 6628 | 143 * Clean up functions in input.cc that handle user input (there |
| 144 currently seems to be some unnecessary duplication of code and it | |
| 145 seems overly complex). | |
| 2330 | 146 |
| 147 * Consider allowing an arbitrary property list to be attached to any | |
| 148 variable. This could be a more general way to handle the help | |
| 149 string that can currently be added with `document'. | |
| 150 | |
| 151 * Allow more command line options to be accessible as built-in | |
| 152 variables (--echo-commands, etc.). | |
| 153 | |
| 154 * Make the interpreter run faster. | |
| 155 | |
| 4325 | 156 * Allow arbitrary lower bounds for array indexing. |
| 2330 | 157 |
| 4325 | 158 * Improve performance of recursive function calls. |
| 2330 | 159 |
| 160 * Improve the way ignore_function_time_stamp works to allow | |
| 3167 | 161 selecting by individual directories or functions. |
| 2330 | 162 |
| 163 * Add a command-line option to tell Octave to just do syntax | |
| 164 checking and not execute statements. | |
| 165 | |
| 166 * Clean up symtab and variable stuff. | |
| 167 | |
| 168 * Input stream class for parser files -- must manage buffers for | |
| 169 flex and context for global variable settings. | |
| 170 | |
| 3162 | 171 * make parser do more semantic checking, continue after errors when |
| 172 compiling functions, etc. | |
| 173 | |
| 2330 | 174 * Make LEXICAL_ERROR have a value that is the error message for |
| 175 parse_error() to print? | |
| 176 | |
| 177 * Add a run-time alias mechanism that would allow things like | |
| 178 | |
| 179 alias fun function_with_a_very_long_name | |
| 180 | |
| 181 so that `function_with_a_very_long_name' could be invoked as | |
| 182 `fun'. | |
| 183 | |
| 184 * Allow local changes to variables to be written more compactly than | |
| 185 is currently possible with unwind_protect. For example, | |
| 186 | |
| 187 function f () | |
| 188 local prefer_column_vectors = something; | |
| 189 ... | |
| 190 endfunction | |
| 191 | |
| 192 would be equivalent to | |
| 193 | |
| 194 function f () | |
| 3758 | 195 save_prefer_column_vectors = prefer_column_vectors; |
| 2330 | 196 unwind_protect |
| 197 prefer_column_vectors = something; | |
| 198 ... | |
| 199 unwind_protect_cleanup | |
| 200 prefer_column_vectors = save_prefer_column_vectors; | |
| 201 end_unwind_protect | |
| 202 endfunction | |
| 203 | |
| 204 * Fix all function files to check for bogus inputs (wrong number or | |
| 205 types of input arguments, wrong number of output arguments). | |
| 206 | |
| 207 * Handle options for built-in functions more consistently. | |
| 208 | |
| 209 * Too much time is spent allocating and freeing memory. What can be | |
| 210 done to improve performance? | |
| 211 | |
| 212 * Error output from Fortran code is ugly. Something should be done to | |
| 213 make it look better. | |
| 214 | |
| 215 * It would be nice if output from the Fortran routines could be | |
| 216 passed through the pager. | |
| 217 | |
| 218 * Attempt to recognize common subexpressions in the parser. | |
| 219 | |
| 220 * Consider making it possible to specify an empty matrix with a | |
| 221 syntax like [](e1, e2). Of course at least one of the expressions | |
| 222 must be zero... | |
| 223 | |
| 224 * Is Matrix::fortran_vec() really necessary? | |
| 2862 | 225 |
| 2746 | 226 * Rewrite whos and the symbol_record_info class. Write a built-in |
| 227 function that gives all the basic information, then write who and | |
| 228 whos as M-files. | |
| 2439 | 229 |
| 2799 | 230 * On systems that support matherr(), make it possible for users to |
| 231 enable the printing of warning messages. | |
| 232 | |
| 2862 | 233 * Make it possible to mark variables and functions as read-only. |
| 234 | |
| 3092 | 235 * Make it possible to write a function that gets a reference to a |
| 236 matrix in memory and change one or more elements without | |
| 237 generating a second copy of the data. | |
| 238 | |
| 5228 | 239 * Use nanosleep instead of usleep if it is available? Apparently |
| 240 nanosleep is to be preferred over usleep on Solaris systems. | |
| 241 | |
| 9213 | 242 -------- |
| 243 Graphics: | |
| 244 -------- | |
| 245 | |
| 246 * Correctly handle case where DISPLAY is unset. Provide | |
| 247 --no-window-system or --nodisplay (?) option. Provide | |
| 248 --display=DISPLAY option? How will this work with gnuplot (i.e., | |
| 249 how do we know whether gnuplot requires an X display to display | |
| 250 graphics)? | |
| 251 | |
| 2330 | 252 ------- |
| 253 History: | |
| 254 ------- | |
| 255 | |
| 256 * Add an option to allow saving input from script files in the | |
| 257 history list. | |
| 258 | |
| 3092 | 259 * The history command should accept two numeric arguments to |
| 260 indicate a range of history entries to display, save or read. | |
| 261 | |
| 5026 | 262 * Avoid writing the history file if the history list has not |
| 263 changed. | |
| 264 | |
| 265 * Avoid permission errors if the history file cannot be opened for | |
| 266 writing. | |
| 267 | |
| 2330 | 268 * Fix history problems -- core dump if multiple processes are |
| 269 writing to the same history file? | |
| 270 | |
| 271 ------------------------------ | |
| 272 Configuration and Installation: | |
| 273 ------------------------------ | |
| 274 | |
| 8921 | 275 * Split config.h into a part for Octave-specific configuration |
| 276 things (this part can be installed) and the generic HAVE_X type of | |
| 277 configure information that should not be installed. | |
| 2473 | 278 |
| 2330 | 279 * Makefile changes: |
| 280 -- eliminate for loops | |
| 281 -- define shell commands or eliminate them | |
| 282 -- consolidate targets | |
| 283 | |
| 284 * Make it possible to configure so that installed binaries and | |
| 285 shared libraries are stripped. | |
| 286 | |
| 3069 | 287 * Create a docs-only distribution? |
| 288 | |
| 2330 | 289 ------------------------------ |
| 290 Documentation and On-Line Help: | |
| 291 ------------------------------ | |
| 292 | |
| 293 * Document new features. | |
| 294 | |
| 295 * Improve the Texinfo Documentation for the interpreter. It would | |
| 296 be useful to have lots more examples, to not have so many forward | |
| 297 references, and to not have very many simple lists of functions. | |
| 298 | |
| 299 * The docs should mention something about efficiency and that using | |
| 300 array operations is almost always a good idea for speed. | |
| 301 | |
| 302 * Texinfo documentation for the C++ classes. | |
| 303 | |
| 304 * Make index entries more consistent to improve behavior of `help -i'. | |
| 305 | |
| 306 * Make `help -i' try to find a whole word match first. | |
| 307 | |
| 3130 | 308 * Clean up help stuff. |
| 309 | |
| 2330 | 310 * Demo files. |
| 311 | |
| 312 ----- | |
| 313 Tests: | |
| 314 ----- | |
| 315 | |
| 316 * Improved set of tests: | |
| 317 | |
| 318 -- Tests for various functions. Would be nice to have a test file | |
| 319 corresponding to every function. | |
| 320 | |
| 321 -- Tests for element by element operators: | |
| 322 + - .* ./ .\ .^ | & < <= == >= > != ! | |
| 323 | |
| 324 -- Tests for boolean operators: && || | |
| 325 | |
| 326 -- Tests for other operators: * / \ ' .' | |
| 327 | |
| 328 -- Tests from bug reports. | |
| 329 | |
| 330 -- Tests for indexed assignment. Need to consider the following: | |
| 331 o fortran-style indexing | |
| 332 o zero-one indexing | |
| 333 o assignment of empty matrix as well as values | |
| 334 o resizing | |
| 335 | |
| 336 * Tests for all internal functions. | |
| 337 | |
| 338 ----------- | |
| 339 Programming: | |
| 340 ----------- | |
| 341 | |
| 8921 | 342 * C++ namespace for Octave library functions. |
| 343 | |
| 2475 | 344 * Better error messages for missing operators? |
| 345 | |
| 346 * Eliminate duplicate enums in pt-exp.cc, pt-const.cc, and ov.cc. | |
| 347 | |
| 348 * Handle octave_print_internal() stuff at the liboctave level. Then | |
| 349 the octave_value classes could just call on the print() methods | |
| 350 for the underlying classes. | |
| 351 | |
| 352 * As much as possible, eliminate explicit checks for the types of | |
| 353 octave_value objects so that user-defined types will automatically | |
| 354 do the right thing in more cases. | |
| 355 | |
| 2330 | 356 * Only include config.h in files that actually need it, instead of |
| 357 including it in every .cc file. Unfortunately, this might not be | |
| 358 so easy to figure out. | |
| 359 | |
| 360 * GNU coding standards: | |
| 361 | |
| 362 -- Add a `Makefile' target to the Makefiles. | |
| 363 -- Comments on #else and #endif preprocessor commands. | |
| 364 -- Change error message format to match standards everywhere. | |
| 365 | |
| 366 * Eliminate more global variables. | |
| 367 | |
| 368 * Move procstream to liboctave. | |
| 369 | |
| 370 * Use references and classes in more places. | |
| 371 | |
| 372 * Share more code among the various *_options functions. | |
| 373 | |
| 374 ------------- | |
| 375 Miscellaneous: | |
| 376 ------------- | |
| 377 | |
| 378 * Implement some functions for interprocess communication: bind, | |
| 379 accept, connect, gethostbyname, etc. | |
| 380 | |
| 2454 | 381 * The installation process should also install octave.el. This |
| 382 needs to detect the appropriate Emacs binary to use to | |
| 383 byte-compile the .el file. Following GNU Emacs philosophy, | |
| 384 installation would be into $(prefix)/share/emacs/site-lisp by | |
| 385 default, but it should be selectable. | |
| 386 | |
| 2330 | 387 * The ability to transparently handle very large files: |
| 388 | |
| 389 Juhana K Kouhia <kouhia@nic.funet.fi> wrote: | |
| 390 | |
| 391 If I have a one-dimensional signal data with the size 400 | |
| 392 Mbytes, then what are my choices to operate with it: | |
| 393 | |
| 394 * I have to split the data | |
| 395 * Octave has a virtual memory on its own and I don't have to | |
| 396 worry about the splitting. | |
| 397 | |
| 398 If I split the data, then my easily programmed processing | |
| 399 programs will become hard to program. | |
| 400 | |
| 401 If possible, I would like to have the virtual memory system in | |
|
9031
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
402 Octave i.e., the all big files, the user see as one big array or |
| 2330 | 403 such. There could be several user selectable models to do the |
| 404 virtual memory depending on what kind of data the user have (1d, | |
| 405 2d) and in what order they are processed (stream or random | |
| 406 access). | |
| 407 | |
| 408 Perhaps this can be done entirely with a library of M-files. | |
| 409 | |
| 3136 | 410 * An interface to gdb. |
| 411 | |
| 412 Michael Smolsky <fnsiguc@weizmann.weizmann.ac.il> wrote: | |
| 413 | |
| 414 I was thinking about a tool, which could be very useful for me | |
|
9031
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
415 in my numerical simulation work. It is an interconnection |
|
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
416 between gdb and octave. We are often managing very large arrays |
| 3136 | 417 of data in our fortran or c codes, which might be studied with |
|
9031
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
418 the help of octave at the algorithm development stages. Assume |
| 3136 | 419 you're coding, say, wave equation. And want to debug the |
|
9031
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
420 code. It would be great to pick some array from the memory of |
|
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
421 the code you're developing, fft it and see the image as a |
|
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
422 log-log plot of the spectral density. I'm facing similar |
| 3136 | 423 problems now. To avoid high c-development cost, I develop in |
|
9031
1052a66078cf
Documentation cleanup of top-level Octave directory (READMEs, INSTALL)
Rik <rdrider0-list@yahoo.com>
parents:
8921
diff
changeset
|
424 matlab/octave, and then rewrite into c. It might be so much |
| 3136 | 425 easier, if I could off-load a c array right from the debugger |
| 426 into octave, study it, and, perhaps, change some [many] values | |
| 427 with a convenient matlab/octave syntax, similar to | |
| 428 a(:,50:250)=zeros(100,200), and then store it back into the | |
| 429 memory of my c code. | |
| 430 | |
| 2789 | 431 * Add a definition to lgrind so that it supports Octave. |
| 432 (See http://www.tex.ac.uk/tex-archive/support/lgrind/ for more | |
| 433 information about lgrind.) | |
| 434 | |
| 2330 | 435 ------ |
| 436 Always: | |
| 437 ------ | |
| 438 | |
| 439 * Squash bugs. | |
| 440 | |
| 441 --30-- | |
| 3162 | 442 </pre> |
| 443 </html> |