Re: Incorrect behaviour for large s-expressions :(
JVMs has a strange limitation for the size of methods. I don't know if there is a plan to solve that on the JVM side. It is quite hard to solve that on the compiler side. When I bumped inot this (once for an ICFP contest), I rewrote a macro so that it emitted smaller methods called from a big method. On Sun, Nov 14, 2010 at 5:48 AM, Robert McIntyre r...@mit.edu wrote: So my friend and I were screwing around, battling versions of LISP as nerds are wont to do, when I came across this: (eval `(clojure.core/+ ~@(take 1e4 (iterate inc 1 Invalid method Code length 89884 in class file user$eval13607 This is just trying to evaluate + directly on a bunch of arguments. Common Lisp on my friend's 30 year old Lisp machine does the equivalent of this with ease, even for much larger numbers. As I'm writing this, my friend is rubbing in this in my face by also doing the above with C-LISP on his laptop. (although his stack overflows for 1e5) I'm losing my battle!!! :( Pls. help! --Robert McIntyre -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- Sent from an IBM Model M, 15 August 1989. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
On Sun, 14 Nov 2010 00:48:13 -0500 Robert McIntyre r...@mit.edu wrote: So my friend and I were screwing around, battling versions of LISP as nerds are wont to do, when I came across this: (eval `(clojure.core/+ ~@(take 1e4 (iterate inc 1 Invalid method Code length 89884 in class file user$eval13607 This is just trying to evaluate + directly on a bunch of arguments. I'd say the first problem is using the macro-building constructs outside a macro. I believe this is generally a bad idea. If you build the list and apply + to it directly, it works fine Clojure 1.2.0 user= (apply + (doall (take 1e4 (iterate inc 1 (apply + (take 1e4 (iterate inc 1))) 50005000 user= (apply + (doall (take 1e5 (iterate inc 1 (apply + (take 1e5 (iterate inc 1))) 55 user= But nope, you've got a real problem. It appears to be with eval: user= (eval (cons + (take 1e4 (iterate inc 1 (eval (cons + (take 1e4 (iterate inc 1 java.lang.ClassFormatError: Invalid method Code length 89881 in class file user$eval26 (NO_SOURCE_FILE:8) user= (count (cons + (take 1e4 (iterate inc 1 (count (cons + (take 1e4 (iterate inc 1 10001 Of course, eval isn't idiomatic clojure. Common Lisp on my friend's 30 year old Lisp machine does the equivalent of this with ease, even for much larger numbers. As I'm writing this, my friend is rubbing in this in my face by also doing the above with C-LISP on his laptop. (although his stack overflows for 1e5) Well, the apply version works out to 1e8 for me if I leave out the doall. If I use the doall, it runs out of heap at 1e7. I'm a little surprised that they're different - I figured apply would instantiate the sequence, and I'd need to use reduce instead of apply for really large sequences. mike -- Mike Meyer m...@mired.org http://www.mired.org/consulting.html Independent Network/Unix/Perforce consultant, email for more information. O ascii ribbon campaign - stop html mail - www.asciiribbon.org -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Should Math/abs be able to accept Ratios?
Math/abs won't accept ratios, so I assumed the Java Math functions only took base number types but all the other methods I tried accept ratios just fine. user= (Math/sin 1/2) 0.479425538604203 user= (Math/sqrt 1/2) 0.7071067811865476 user= (Math/pow 1/2 1/2) 0.7071067811865476 user= (Math/abs 1/2) IllegalArgumentException No matching method found: abs clojure.lang.Reflector.invokeMatchingMethod (Reflector.java:77) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
expand question
(map vector [1 2 3] [4 5 6] [7 8 9]) gives back ([1 4 7] [2 5 8] [3 6 9]) but if I have a function that returns the list: '([1 2 3] [4 5 6] [7 8 9]) and call map vector with the list, then I no longer get ([1 4 7] [2 5 8] [3 6 9]) i.e. (def x '([1 2 3] [4 5 6] [7 8 9])) (map vector x) ==([[1 2 3]] [[4 5 6]] [[7 8 9]]) so is there a way to effectively expand the argument x (map vector x) to be (map vector [1 2 3] [4 5 6] [7 8 9]) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: expand question
(apply (partial map vector) [[1 2 3] [4 5 6] [7 8 9]]) ([1 4 7] [2 5 8] [3 6 9]) On Sun, Nov 14, 2010 at 2:16 PM, garf gary.overg...@gmail.com wrote: (map vector [1 2 3] [4 5 6] [7 8 9]) gives back ([1 4 7] [2 5 8] [3 6 9]) but if I have a function that returns the list: '([1 2 3] [4 5 6] [7 8 9]) and call map vector with the list, then I no longer get ([1 4 7] [2 5 8] [3 6 9]) i.e. (def x '([1 2 3] [4 5 6] [7 8 9])) (map vector x) ==([[1 2 3]] [[4 5 6]] [[7 8 9]]) so is there a way to effectively expand the argument x (map vector x) to be (map vector [1 2 3] [4 5 6] [7 8 9]) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.comclojure%2bunsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- Moritz Ulrich Programmer, Student, Almost normal Guy http://www.google.com/profiles/ulrich.moritz BB5F086F-C798-41D5-B742-494C1E9677E8 -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
@Nicolas Oury I'd normally agree that you just shouldn't do that kind of thing, but this is not correct behaviour for something that I would consider a basic thing. It feels dirty that our functions don't work for arbitrary arities but instead for arities only between 0 and around 7000. Since apply does work for this, would it be possible for the reader to convert such long forms into something like (apply + (list long-list)) before evaluating them? @Mike Meyer Using apply is different than what I'm doing. When I use eval I'm trying to evaluate a huge s-expression. When you use apply you're evaluating a s-expression with three elements. Same thing with the count form (except with two elements). The problem isn't because I'm calling eval or not using idiomatic clojure; I just wrote it that way so it would only take one line. This version doesn't use anything non-idiomatic and still gives the same incorrect behavior: (clojure.core/+ 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031
Re: expand question
The use of partial is unnecessary because apply takes any number of arguments and expands its last argument. (apply map vector [[1 2 3] [4 5 6] [7 8 9]]) is equivalent to (map vector [1 2 3] [4 5 6] [7 8 9]) and results in ([1 4 7] [2 5 8] [3 6 9]) On Sun, Nov 14, 2010 at 8:30 AM, Moritz Ulrich ulrich.mor...@googlemail.com wrote: (apply (partial map vector) [[1 2 3] [4 5 6] [7 8 9]]) ([1 4 7] [2 5 8] [3 6 9]) On Sun, Nov 14, 2010 at 2:16 PM, garf gary.overg...@gmail.com wrote: (map vector [1 2 3] [4 5 6] [7 8 9]) gives back ([1 4 7] [2 5 8] [3 6 9]) but if I have a function that returns the list: '([1 2 3] [4 5 6] [7 8 9]) and call map vector with the list, then I no longer get ([1 4 7] [2 5 8] [3 6 9]) i.e. (def x '([1 2 3] [4 5 6] [7 8 9])) (map vector x) ==([[1 2 3]] [[4 5 6]] [[7 8 9]]) so is there a way to effectively expand the argument x (map vector x) to be (map vector [1 2 3] [4 5 6] [7 8 9]) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- Moritz Ulrich Programmer, Student, Almost normal Guy http://www.google.com/profiles/ulrich.moritz BB5F086F-C798-41D5-B742-494C1E9677E8 -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
On 14 November 2010 15:43, Robert McIntyre r...@mit.edu wrote: @Nicolas Oury I'd normally agree that you just shouldn't do that kind of thing, but this is not correct behaviour for something that I would consider a basic thing. It feels dirty that our functions don't work for arbitrary arities but instead for arities only between 0 and around 7000. Since apply does work for this, would it be possible for the reader to convert such long forms into something like (apply + (list long-list)) before evaluating them? I believe the underlying problem is a limit of the JVM. Maybe it would be possible for the Clojure compiler to work around the limitation, though. -- Michael Wood esiot...@gmail.com -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Should Math/abs be able to accept Ratios?
There are four separate methods called Math/abs, to handle the following types: int, long, float, double. So when you use Math/abs on a different Number type, it is not clear which of those methods it should use. The other examples you gave can only accept double. Maybe in that case Clojure is automatically converting to double. It seems like it would be an improvement if ratios had a preference order for what they could be cast into, in which double is preferable to float is preferable to long is preferable to int. On the other hand, java.lang.Math should really only be used when you want fast primitive operations. If you want better support for Clojure numeric types, try clojure.contrib.math. http://richhickey.github.com/clojure-contrib/math-api.html On Sun, Nov 14, 2010 at 6:46 AM, Jarl Haggerty fictivela...@gmail.com wrote: Math/abs won't accept ratios, so I assumed the Java Math functions only took base number types but all the other methods I tried accept ratios just fine. user= (Math/sin 1/2) 0.479425538604203 user= (Math/sqrt 1/2) 0.7071067811865476 user= (Math/pow 1/2 1/2) 0.7071067811865476 user= (Math/abs 1/2) IllegalArgumentException No matching method found: abs clojure.lang.Reflector.invokeMatchingMethod (Reflector.java:77) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
I believe the underlying problem is a limit of the JVM. Maybe it would be possible for the Clojure compiler to work around the limitation, though. It has a non trivial interaction with the lack of TCO. The trivial transformation of going from a big method to a sequence of small methods could blow the stack. You could maybe do a tree of function call and emit a warning though. Does someone know what Scala does? Nicolas. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: expand question
How about just user= (vector [1 2 3] [4 5 6] [7 8 9]) [[1 2 3] [4 5 6] [7 8 9]] user= -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
On Sun, 14 Nov 2010 08:43:11 -0500 Robert McIntyre r...@mit.edu wrote: @Mike Meyer Using apply is different than what I'm doing. Yup. When I use eval I'm trying to evaluate a huge s-expression. When you use apply you're evaluating a s-expression with three elements. Same thing with the count form (except with two elements). The problem isn't because I'm calling eval or not using idiomatic clojure; I just wrote it that way so it would only take one line. I did agree that there was a problem. The thing is, quasiquotes in clojure were designed for use in macros, and using them outside macros sometimes generates weird results: I wanted to make sure that wasn't the case here. My first attempt - in idiomatic clojure - didn't recreate it. So I went a bit further afield to do so. Are we really OK with having a 30 year old (Common Lisp/Lisp Machine) that operates at megahertz speeds do better than (clojure/JVM) here? Yes, I'm OK that a LISP running on an architecture that's the end result of decades of research on creating machines that run LISP well has fewer and/or higher limits than a LISP running on a VM designed to run Java. I'm not even sure it's worth any effort in fixing. You're not going to run into this limit except in machine-generated code, and there's an easy work-around: generate (apply fun (sequence)) instead of (fun sequence). mike -- Mike Meyer m...@mired.org http://www.mired.org/consulting.html Independent Network/Unix/Perforce consultant, email for more information. O ascii ribbon campaign - stop html mail - www.asciiribbon.org -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: ANN: Programothesis screencast series on Clojure, Emacs, etc
On Fri, Nov 12, 2010 at 4:25 PM, Scott Jaderholm jaderh...@gmail.com wrote: I don't think there are any great new movies in the theater this weekend so if you're looking to kick back and relax and watch the tube a bit you might checkout the first few episodes of my new screencast series on Clojure, Emacs, Slime, etc. http://youtube.com/emailataskcom The episodes were recorded several months ago but just posted today. I expect to post a ton more real soon so subscribe if you want to be notified. Thanks! These look quite cool! -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Performance of seq on empty collections
In the API it is suggested to use `seq` to check if coll is empty. I was working on some code recently found that my biggest performance bottleneck was calling `seq` to check for emptiness. The calls to `seq` were causing lots of object allocation and taking noticeable CPU time. I switched to using `identical?` to explicitly compare against the empty vector and was rewarded with a drastic reduction in execution time. Here are some hasty tests showing just how big the difference can be: user= (let [iterations 1] (time (dotimes [_ iterations] (identical? [] []))) (time (dotimes [_ iterations] (seq [] Elapsed time: 3.512 msecs Elapsed time: 2512.366 msecs nil user= (let [iterations 1] (time (dotimes [_ iterations] (identical? ))) (time (dotimes [_ iterations] (seq Elapsed time: 3.898 msecs Elapsed time: 5607.865 msecs nil user= (let [iterations 1] (time (dotimes [_ iterations] (identical? () ( (time (dotimes [_ iterations] (seq () Elapsed time: 3.768 msecs Elapsed time: 2258.095 msecs nil Has any thought been given to providing a faster `empty?` that is not based on seq? Thanks, Eric Kobrin -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
On Sun, Nov 14, 2010 at 2:16 PM, Eric Kobrin erl...@gmail.com wrote: In the API it is suggested to use `seq` to check if coll is empty. I was working on some code recently found that my biggest performance bottleneck was calling `seq` to check for emptiness. The calls to `seq` were causing lots of object allocation and taking noticeable CPU time. I switched to using `identical?` to explicitly compare against the empty vector and was rewarded with a drastic reduction in execution time. Only there's a wee problem: user= (identical? [] (pop [3])) false Not all empty vectors are considered identical. = works: user= (= [] (pop [3])) true = is also ten times SLOWER than seq if the empty vectors are not identical, though it's as fast as identical? when they are: user= (let [iterations 1] (time (dotimes [_ iterations] (= [] (pop [3] (time (dotimes [_ iterations] (seq [] Elapsed time: 29994.93852 msecs Elapsed time: 2745.21924 msecs This is troubling. The seq function should be faster than this, and = on nonidentical empty colls should be WAY faster than this. On the other hand, the identical? test taking 4ms for 100,000,000 iterations suggests it was optimized away to nothing by the JIT compiler. Unless of course both of us have 100GHz CPUs that can execute a fresh instruction every 10 picoseconds. :) -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
On Nov 14, 2010, at 2:16 PM, Eric Kobrin wrote: In the API it is suggested to use `seq` to check if coll is empty. Your timing results raise some interesting questions, however, the API doesn't suggest using 'seq' to check if a collection is empty. That's what 'empty?' is for. The documentation note suggests (for style purposes apparently) that you use 'seq' to test that the collection is not empty. So to be precise you are testing two different things below. For instance, (identical? coll []) is true when coll is an empty vector. (seq coll) is true when coll is not empty. The correct equivalent would be to test (empty? coll). Of course, this doesn't change the results. I get similar timings with empty?: user= (let [iterations 1] (time (dotimes [_ iterations] (identical? [] []))) (time (dotimes [_ iterations] (empty? [] Elapsed time: 2.294 msecs Elapsed time: 2191.256 msecs nil user= (let [iterations 1] (time (dotimes [_ iterations] (identical? ))) (time (dotimes [_ iterations] (empty? Elapsed time: 2.657 msecs Elapsed time: 4654.622 msecs nil user= (let [iterations 1] (time (dotimes [_ iterations] (identical? () ( (time (dotimes [_ iterations] (empty? () Elapsed time: 2.608 msecs Elapsed time: 2144.142 msecs nil This isn't so surprising though, considering that 'identical?' is the simplest possible test you could try--do two references point to the same object in memory? It can't get any more efficient than that. Have all good days, David Sletten I was working on some code recently found that my biggest performance bottleneck was calling `seq` to check for emptiness. The calls to `seq` were causing lots of object allocation and taking noticeable CPU time. I switched to using `identical?` to explicitly compare against the empty vector and was rewarded with a drastic reduction in execution time. Here are some hasty tests showing just how big the difference can be: user= (let [iterations 1] (time (dotimes [_ iterations] (identical? [] []))) (time (dotimes [_ iterations] (seq [] Elapsed time: 3.512 msecs Elapsed time: 2512.366 msecs nil user= (let [iterations 1] (time (dotimes [_ iterations] (identical? ))) (time (dotimes [_ iterations] (seq Elapsed time: 3.898 msecs Elapsed time: 5607.865 msecs nil user= (let [iterations 1] (time (dotimes [_ iterations] (identical? () ( (time (dotimes [_ iterations] (seq () Elapsed time: 3.768 msecs Elapsed time: 2258.095 msecs nil Has any thought been given to providing a faster `empty?` that is not based on seq? Thanks, Eric Kobrin -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
On Sun, Nov 14, 2010 at 2:42 PM, Ken Wesson kwess...@gmail.com wrote: user= (let [iterations 1] (time (dotimes [_ iterations] (= [] (pop [3] (time (dotimes [_ iterations] (seq [] Elapsed time: 29994.93852 msecs Elapsed time: 2745.21924 msecs It occurred to me that the JIT might not be hoisting the constant (pop [3]) out of the loop so I retested with this: user= (let [iterations 1 ev (pop [3])] (time (dotimes [_ iterations] (= [] ev))) (time (dotimes [_ iterations] (seq [] Elapsed time: 16477.68672 msecs Elapsed time: 2105.01128 msecs Still nearly 10x slower. The JIT did something (optimized the function call overhead away, most likely) but it didn't hoist the entire thing, and most of the time is indeed spent in the = test. = is damned slow. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: shorter alternatives for `comp' and `partial'
On Sat, Nov 13, 2010 at 11:33 PM, Eric Schulte schulte.e...@gmail.com wrote: I'm about to begin starting all of my clojure namespaces with (def o comp) ; o for cOmp, Haskell's (.) or Mathematical composition \circ (def p partial) ; p for partial snip Is there any support for including these function aliases for `comp' and `partial' (or some other shortened names) in the core? Are you really sure you won't want to use 'p' to represent something like a probability, at some point? I suspect, the shorter your name, the more likely you are to have a naming clash. That, and you're likely to have some problems remembering what it's for. Clojure code (and lisp in general) tends to be rather small for what it does, anyway. So far, we seem to have avoided the emacs-style-long-hyphenated-sentence function names. If you're playing golf, do whatever, but for Clojure's core? I'd personally prefer 'comp' and 'partial', myself. -- Chris Riddoch -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
LabREPL status ... github returning 404's
Hi all, When I ran some Clojure Dojo's in Dundee, we did a few group sessions going through labrepl and its exercises; and I thought the format worked well. I'm keen to do the same thing again, here in Manchester (UK): http://madlab.org.uk/content/kick-ass-coding-with-clojure/ Unfortunately the labrepl github page is now returning 404's. Is Labrepl still active, or is github playing up? Can anyone advise me on where to get the latest version? R. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
In the particular bit of code I'm working on, I have a collection of vectors. The last one is always [] and no others are empty. Using `identical?` instead of `seq` to detect that last vector shaved quite a bit of time in my loop. This brought to mind the general case of detecting emptiness. The current practice of using `seq` to check for non-emptiness wastes resources. This limits the use of nice abstractions like `reduce` in performance critical environments where a sentinel could be used. Not everyone will be able to use a sentinel value to solve their particular problem. What can be done to make `empty` faster? -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: LabREPL status ... github returning 404's
Github is down at the moment unfortunately... On Sun, Nov 14, 2010 at 3:27 PM, Rick Moynihan rick.moyni...@gmail.comwrote: Hi all, When I ran some Clojure Dojo's in Dundee, we did a few group sessions going through labrepl and its exercises; and I thought the format worked well. I'm keen to do the same thing again, here in Manchester (UK): http://madlab.org.uk/content/kick-ass-coding-with-clojure/ Unfortunately the labrepl github page is now returning 404's. Is Labrepl still active, or is github playing up? Can anyone advise me on where to get the latest version? R. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.comclojure%2bunsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
That is not in fact an adequate workaround --- (eval `(apply + ~@(take 9001 (iterate inc 1 ;; OVER 9000!!! or, alternately (eval (cons 'apply (cons '+ (take 9001 (iterate inc 1) will fail just as in the addition examples. It's not true that you can just use an apply in your auto generated code, you would instead have to do something like a tree of function calls, so It may be worth increasing the limit for for the sake of enabling machine generated code. What are you peoples' thoughts on this? --Robert McIntyre On Sun, Nov 14, 2010 at 12:54 PM, Mike Meyer mwm-keyword-googlegroups.620...@mired.org wrote: On Sun, 14 Nov 2010 08:43:11 -0500 Robert McIntyre r...@mit.edu wrote: @Mike Meyer Using apply is different than what I'm doing. Yup. When I use eval I'm trying to evaluate a huge s-expression. When you use apply you're evaluating a s-expression with three elements. Same thing with the count form (except with two elements). The problem isn't because I'm calling eval or not using idiomatic clojure; I just wrote it that way so it would only take one line. I did agree that there was a problem. The thing is, quasiquotes in clojure were designed for use in macros, and using them outside macros sometimes generates weird results: I wanted to make sure that wasn't the case here. My first attempt - in idiomatic clojure - didn't recreate it. So I went a bit further afield to do so. Are we really OK with having a 30 year old (Common Lisp/Lisp Machine) that operates at megahertz speeds do better than (clojure/JVM) here? Yes, I'm OK that a LISP running on an architecture that's the end result of decades of research on creating machines that run LISP well has fewer and/or higher limits than a LISP running on a VM designed to run Java. I'm not even sure it's worth any effort in fixing. You're not going to run into this limit except in machine-generated code, and there's an easy work-around: generate (apply fun (sequence)) instead of (fun sequence). mike -- Mike Meyer m...@mired.org http://www.mired.org/consulting.html Independent Network/Unix/Perforce consultant, email for more information. O ascii ribbon campaign - stop html mail - www.asciiribbon.org -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
On Sun, Nov 14, 2010 at 3:36 PM, Eric Kobrin erl...@gmail.com wrote: In the particular bit of code I'm working on, I have a collection of vectors. The last one is always [] and no others are empty. Using `identical?` instead of `seq` to detect that last vector shaved quite a bit of time in my loop. The use of identical? is just wrong. It checks whether two objects are in fact the same object in memory. The fact that: (identical? [] []) ; true is just an implementation detail. The compiler *happened* to use the same object - it's an optimization for constant literals. Better to use a real sentinel. Also such low timings for 1e8 iterations should being a warning sign that perhaps you are not measuring what you think you are measuring. Your (identical? [] []) is likely becoming a no-op. In the general case, if you're using vectors and loop anyway for performance reasons, (= (count v) 0) to check for a empty vector is plenty fast and resource efficient. David -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
On Sun, Nov 14, 2010 at 4:33 PM, David Nolen dnolen.li...@gmail.com wrote: On Sun, Nov 14, 2010 at 3:36 PM, Eric Kobrin erl...@gmail.com wrote: In the particular bit of code I'm working on, I have a collection of vectors. The last one is always [] and no others are empty. Using `identical?` instead of `seq` to detect that last vector shaved quite a bit of time in my loop. The use of identical? is just wrong. It checks whether two objects are in fact the same object in memory. The fact that: (identical? [] []) ; true is just an implementation detail. The compiler *happened* to use the same object - it's an optimization for constant literals. Better to use a real sentinel. Also such low timings for 1e8 iterations should being a warning sign that perhaps you are not measuring what you think you are measuring. Your (identical? [] []) is likely becoming a no-op. In the general case, if you're using vectors and loop anyway for performance reasons, (= (count v) 0) to check for a empty vector is plenty fast and resource efficient. Interestingly, (= (count v) 0) is relatively fast (900 msec); (seq v) is slower (2s); (empty? v) is even slower (3s); and (= v []) is slowest (16s). As for a real sentinel, nothing beats the old (Object.) -- it's guaranteed not equal to any existing object at time of creation, so no preexisting piece of data in your input will look like a freshly-created one, unlike the case with such traditional sentinels as nil and zero. (The zero sentinel in C strings has caused no end of trouble whenever the input string could have an ASCII NUL in it, and we're all sick to death of Java collections that choke on null elements, keys, or values. ConcurrentHashMap is a notorious culprit there.) (let [sentinel (Object.) v (conj some-vector sentinel)] (loop [...] (if (= sentinel something) ... Another nice thing about (Object.) is that the Object .equals method boils down to the Java == operator, and JIT-optimizes away to same, so the above is just as fast as, and more readable than, (if (identical? sentinel something) ...). Just be sure to put the sentinel first so the object whose method is being called is an unspecialized Object. If it's the other way around something else's .equals gets called, and if it's well designed it will start with if (other == this) return true;, and branch prediction on the JITted code will eventually eliminate the cost of the rest of the method not just being return false;, but you never know. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
On Nov 14, 2010, at 4:04 PM, Ken Wesson wrote: Interestingly, (= (count v) 0) is relatively fast (900 msec); (seq v) is slower (2s); (empty? v) is even slower (3s); and (= v []) is slowest (16s). Strange. Here are the timings I get for 1e8 iterations: (zero? (count v)): ~3600 msecs (seq v): ~2300 msecs (empty? v): ~3100 msecs (= v []): ~460 msecs (= 0 (count v)) is even slower than (zero? (count v)). I'm running Clojure 1.2 on a MacBook (Core 2 Duo, 2GHz). -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: shorter alternatives for `comp' and `partial'
Am 14.11.2010 07:33, schrieb Eric Schulte: Hi, I find myself frequently using the `comp' and `partial' functions and while I really enjoy being able to program in a point free style, the length (in characters) of these command names often has the effect of causing what should be a brief statement to span multiple lines. I'm about to begin starting all of my clojure namespaces with (def o comp) ; o for cOmp, Haskell's (.) or Mathematical composition \circ (def p partial) ; p for partial Eric, I think you are approaching it from a wrong way. Shortening those two names won’t increase your productivity, and they will not really improve readability. It takes months or years to write complex software. Having o and p won’t help you to reduce such complexity, and as you mentioned yourself — your colleagues might not profit from it too (or future employees). You may however want to checkout the #() reader macro vs. using partial. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Simple Neural Network DSL -- request for feedback
It might be interesting to add a Clojure frontend to Nengo. Having said that, Nengo is very opinionated in the sense that you have to buy in to the kind of neural simulation Eliasmith is doing (and the way he's doing it), which Eric may not be interested in doing. There are many other kinds of neural nets worthy of research besides Eliasmith's NEF. Still, it'd be cool to have a Clojure frontend to Nengo. Carson On Nov 13, 4:09 pm, Ross Gayler r.gay...@gmail.com wrote: You might also consider using your DSL as a frontend to the Nengo neural simulator (http://nengo.ca). Nengo (which is written in Java) has recently added a Python scripting interface (http://www.frontiersin.org/neuroinformatics/ 10.3389/neuro.11/007.2009/abstract). Nengo has a lot to recommend it and is pretty mature, so you may save yourself a lot of effort under the covers - also the way Nengo conceptualises the neyworks might be useful feedback to your DSL design. Ross On Nov 14, 5:18 am, Eric Schulte schulte.e...@gmail.com wrote: Hi Ross, #+begin_src clojure (let [n {:phi identity :accum (comp (partial reduce +) (partial map *)) :weights [2 2 2]}] [(repeat 3 n) (repeat 5 n) (assoc n :weights (vec (repeat 5 1)))]) #+end_src would result in the following connection pattern [[file:/tmp/layers.png]] layers.png 45KViewDownload However, for other NNs you may care about the topological organisation of the neurons in a 1-D, 2-D, or 3-D space in order to do things like connecting corresponding neurons in different layers or having the probability of a connection be a function of the separation of the neurons. In this case, you might use a data structure representing the coordinates of each neuron as its key. Fully agreed, I'm partway through implementing what you've just described (at least as I understand it), in that the library now declares a new Graph data type which consists of a list of keys-Neural mappings as well as a directed edge set. Using this new data type it is possible to construct, run and train arbitrarily connected graphs of Neural elements. See the fourth example athttp://repo.or.cz/w/neural-net.git Best -- Eric Ross- Hide quoted text - - Show quoted text - Ross Gayler r.gay...@gmail.com writes: On Nov 13, 9:12 am, Eric Schulte schulte.e...@gmail.com wrote: Albert Cardona sapri...@gmail.com writes: Your neural network DSL looks great. One minor comment: why use lists instead of sets? ... I used lists because I want to be able to specify a network in which (at least initially) all neurons in a hidden layer are identical e.g. the list example athttp://cs.unm.edu/~eschulte/src/neural-net/. You might want to consider maps. Currently I'm using maps to specify a single neuron, and I fear it would add complexity to have two different meanings for maps. For some NN models all you care about is that each neuron has a unique identity (in which case using an index value as a key is as good a solution as any). I'm currently using lists only for fully connected layers in a neural network, e.g. the following code -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
On Nov 14, 2010, at 4:21 PM, Robert McIntyre wrote: It's not true that you can just use an apply in your auto generated code, you would instead have to do something like a tree of function calls, so It may be worth increasing the limit for for the sake of enabling machine generated code. What are you peoples' thoughts on this? I for one would certainly prefer it if such limits on code size could be avoided. I'm a fan of machine-generated code in many contexts. But if the limit is dictated by the platform, with no straightforward workaround, then I guess we might just have to live with it anyway. -Lee -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Incorrect behaviour for large s-expressions :(
On Sun, Nov 14, 2010 at 4:21 PM, Robert McIntyre r...@mit.edu wrote: That is not in fact an adequate workaround --- (eval `(apply + ~@(take 9001 (iterate inc 1 ;; OVER 9000!!! or, alternately (eval (cons 'apply (cons '+ (take 9001 (iterate inc 1) will fail just as in the addition examples. It's not true that you can just use an apply in your auto generated code, you would instead have to do something like a tree of function calls, so It may be worth increasing the limit for for the sake of enabling machine generated code. What are you peoples' thoughts on this? --Robert McIntyre Not worth getting worked up. David -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: Performance of seq on empty collections
Some more String-specific timings, modified to avoid inlining differences between alternatives: (let [iterations 1e8] (time (dotimes [_ iterations] (= 0 (.length (.substring x 1) (time (dotimes [_ iterations] (seq (.substring x 1) Elapsed time: 3125.125 msecs Elapsed time: 8198.331 msecs (let [iterations 1e8] (time (dotimes [_ iterations] (= 0 (count (.substring x 1) (time (dotimes [_ iterations] (seq (.substring x 1) Elapsed time: 9586.196 msecs Elapsed time: 8006.184 msecs I also tried using String.intern() so that `identical?` could be used, that took longer than any of the other options. Note: I run each of these several times and only include the results once they are consistent between runs. This avoids results tainted by compilation and jit overhead. -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
max-key taking a list rather than different number of args
Hi, I think a better usage for max-key would be (max-key f someseq) rather than passing the values as args. I used (defn max-key-seq [f someseq] (apply max-key (into [f] someseq))) to make it do what I wanted Is there a better way? Cheers, Tom -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: max-key taking a list rather than different number of args
On 15 November 2010 14:17, Tom Hall thattommyh...@gmail.com wrote: Hi, I think a better usage for max-key would be (max-key f someseq) rather than passing the values as args. I used (defn max-key-seq [f someseq] (apply max-key (into [f] someseq))) to make it do what I wanted Is there a better way? You should be able to use (apply max-key f someseq); apply takes a variable number of args, and only the last is expanded. Regards, Stuart -- You received this message because you are subscribed to the Google Groups Clojure group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en