Fri Nov  7 09:10:17 EST 2014  stevenj@alum.mit.edu
  * silence compiler warning

    M ./hcubature.c +1

Fri Nov  7 09:08:21 EST 2014  stevenj@alum.mit.edu
  * fix memory leak in BSD realloc(p, 0); fix stevengj/Cubature.jl#7"

    M ./NEWS +5
    M ./hcubature.c -1 +7

Thu May 23 15:07:44 EDT 2013  stevenj@alum.mit.edu
  * 1.0.1 version bump

    M ./NEWS +4

Thu May 23 15:05:51 EDT 2013  stevenj@alum.mit.edu
  * cubature.h should include stdlib.h to make sure it has the definition of size_t

    M ./cubature.h +2

Fri Mar  1 22:08:46 EST 2013  stevenj@alum.mit.edu
  * add version number 1.0 and NEWS

    M ./Makefile -8 +6
    A ./NEWS

Thu Feb 28 23:13:23 EST 2013  stevenj@alum.mit.edu
  * dylib rules

    M ./Makefile +14

Thu Feb 28 23:05:04 EST 2013  stevenj@alum.mit.edu
  * tarball directory no longer includes date

    M ./Makefile -1 +1

Thu Feb 28 22:49:41 EST 2013  stevenj@alum.mit.edu
  * dll tweak

    M ./Makefile -2 +2

Thu Feb 28 21:21:25 EST 2013  stevenj@alum.mit.edu
  * makefile updates

    M ./Makefile -1 +17

Thu Feb 28 16:07:47 EST 2013  stevenj@alum.mit.edu
  * whoops, fix convergence test for negative integrals

    M ./converged.h -2 +2

Wed Feb 27 17:33:07 EST 2013  stevenj@alum.mit.edu
  * comment fix

    M ./cubature.h -1 +1

Wed Feb 27 17:14:17 EST 2013  stevenj@alum.mit.edu
  * check for integrand error in 0-dimensional case

    M ./hcubature.c -1 +1
    M ./pcubature.c -1 +1

Wed Feb 27 16:51:32 EST 2013  stevenj@alum.mit.edu
  * since I'm changing the API, take advantage and change a few parameters to size_t for 64-bit systems (although it still seems unlikely that > 2^32 points will ever be used for integration)

    M ./README -1 +1
    M ./cubature.h -8 +8
    M ./hcubature.c -17 +19
    M ./pcubature.c -18 +22
    M ./vwrapper.h -1 +1

Wed Feb 27 14:29:34 EST 2013  stevenj@alum.mit.edu
  * increase max clencurt rule (scientific applications can afford a few MB of lookup table, especially since we access it in a cache-friendly way)

    M ./Makefile -1 +1
    M ./README -8 +9

Tue Feb 26 14:31:19 EST 2013  stevenj@alum.mit.edu
  * whoops, fix bug in 1d hcubature introduced by recent re-indexing

    M ./hcubature.c -16 +17

Tue Feb 26 13:49:53 EST 2013  stevenj@alum.mit.edu
  * rename adapt_integrate -> cubature

    M ./README -14 +14
    M ./cubature.h -34 +32
    M ./hcubature.c -24 +24
    M ./pcubature.c -20 +20
    M ./test.c -5 +5

Tue Feb 26 13:48:04 EST 2013  stevenj@alum.mit.edu
  * more updates for release

    M ./Makefile -3 +3
    M ./README -4 +16
    M ./clencurt_gen.c +26
    M ./cubature.h -1 +1
    M ./hcubature.c -1 +1
    M ./pcubature.c +20
    M ./test.c -2 +2
    M ./vwrapper.h -1 +1

Tue Feb 26 13:17:23 EST 2013  stevenj@alum.mit.edu
  * converged.c -> converged.h

     ./converged.c -> ./converged.h
    M ./hcubature.c -1 +1
    M ./pcubature.c -1 +1

Tue Feb 26 13:14:46 EST 2013  stevenj@alum.mit.edu
  * make hcubature and pcubature use the same vectorized format (transpose the former)

    M ./Makefile -2 +2
    M ./README -3 +3
    M ./hcubature.c -47 +25
    M ./pcubature.c -16 +1
    A ./vwrapper.h

Tue Feb 26 12:35:47 EST 2013  stevenj@alum.mit.edu
  * updated docs

    M ./Makefile -4 +4
    M ./README -45 +105
    M ./hcubature.c +1
    M ./pcubature.c +1

Tue Feb 26 10:40:21 EST 2013  stevenj@alum.mit.edu
  * rename adapt_integrate -> hadapt_integrate, support different error norms for vector integrands, support user termination of integration from the integrand function

     ./cubature.c -> ./hcubature.c
    M ./Makefile -8 +5
    A ./converged.c
    M ./cubature.h -28 +46
    M ./hcubature.c -32 +45
    M ./pcubature.c -47 +56
    M ./test.c -7 +8

Wed Feb 22 17:43:43 EST 2012  stevenj@alum.mit.edu
  * whoops, handle > MAXDIM integrands

    M ./pcubature.c -3 +6

Wed Feb 22 16:07:59 EST 2012  stevenj@alum.mit.edu
  * link -lm in pcubature test

    M ./Makefile -2 +4

Wed Feb 22 16:04:24 EST 2012  stevenj@alum.mit.edu
  * allow pcubature_v_buf to take m[] as argument

    M ./cubature.h +1
    M ./pcubature.c -11 +15

Wed Feb 22 15:59:01 EST 2012  stevenj@alum.mit.edu
  * bug fix in pcubature

    M ./pcubature.c -1 +1

Mon Feb 13 13:12:01 EST 2012  stevenj@alum.mit.edu
  * bug fixes in pcubature, now seems to work in >1d

    M ./pcubature.c -6 +6

Sun Feb 12 21:57:31 EST 2012  stevenj@alum.mit.edu
  * initial stab at C-C pcubature (not working in > 1d yet)

    M ./Makefile -2 +16
    A ./clencurt_gen.c
    M ./cubature.h +18
    A ./pcubature.c
    M ./test.c -1 +3

Sun Oct 31 19:35:41 EDT 2010  stevenj@alum.mit.edu
  * one byte per j (4 bits is only enough for j < 16, not 32, duh, and that's too small); another attempt at an error estimate (still not right)

    M ./scubature.c -156 +154

Sun Oct 31 14:36:20 EDT 2010  stevenj@alum.mit.edu
  * turn off VERBOSE, don't evaluate endpoints in singular case

    M ./test.c -4 +10

Sun Oct 31 14:35:29 EDT 2010  stevenj@alum.mit.edu
  * bug fixes in Jp_set and evaluation count

    M ./scubature.c -10 +9

Sun Oct 31 13:54:48 EDT 2010  stevenj@alum.mit.edu
  * eval missing J -- tentatively looking good in > 1d!

    M ./scubature.c -1 +4

Sun Oct 31 13:50:32 EDT 2010  stevenj@alum.mit.edu
  * some bugfixes, > 1d sort of works, but is not yet doing all the J's it should

    M ./scubature.c -5 +5

Sun Oct 31 13:44:57 EDT 2010  stevenj@alum.mit.edu
  * various bugfixes in scubature; 1d case now seems to work, > 1d still untested

    M ./scubature.c -12 +30
    M ./test.c -1 +9

Sun Oct 31 12:27:03 EDT 2010  stevenj@alum.mit.edu
  * separate test program from cubature.c, add scubature API (untested)

    M ./Makefile -3 +11
    M ./cubature.c -268
    M ./cubature.h +7
    M ./scubature.c -36 +108
    A ./test.c

Sun Oct 31 11:40:53 EDT 2010  stevenj@alum.mit.edu
  * always set N = |M|_\infty

    M ./scubature.c -7 +7

Sun Oct 31 11:38:29 EDT 2010  stevenj@alum.mit.edu
  * make redblack.c a header file so users won't try to compile it alone

     ./redblack.c -> ./redblack.h
    M ./scubature.c -1 +1

Sun Oct 31 03:27:38 EDT 2010  stevenj@alum.mit.edu
  * first stab at sparse-grid cubature (not yet finishedd, lonly 1d C-C part is tested)

    A ./redblack.c
    A ./scubature.c

Mon Oct 18 00:35:54 EDT 2010  stevenj@alum.mit.edu
  * updated TODO etc.

    M ./cubature.c -6 +6

Sun Oct 17 22:10:01 EDT 2010  stevenj@alum.mit.edu
  * make sure it compiles in pedantic ANSI mode

    M ./Makefile -1 +1
    M ./cubature.c -6 +10

Sun Oct 17 20:58:49 EDT 2010  stevenj@alum.mit.edu
  * more parallelism in adapt_integrate_v, using an algorithm by Gladstone

    M ./README -1 +19
    M ./cubature.c -88 +152

Sun Oct 17 01:42:50 EDT 2010  stevenj@alum.mit.edu
  * spelling

    M ./cubature.c -1 +1

Sat Oct 16 22:24:45 EDT 2010  stevenj@alum.mit.edu
  * added ChangeLog to dist

    M ./Makefile -2 +7

Sat Oct 16 22:15:22 EDT 2010  stevenj@alum.mit.edu
  * use SUCCESS/FAILURE return codes consistently

    M ./cubature.c -42 +42

Sat Oct 16 21:59:16 EDT 2010  stevenj@alum.mit.edu
  * check for more malloc failures

    M ./cubature.c -19 +22

Sat Oct 16 21:50:10 EDT 2010  stevenj@alum.mit.edu
  * expose a bit more parallelism by evaluating two regions at once (thanks to Alejandro Rodriguez for the suggestion); also, more extensive checking for out-of-memory situations

    M ./cubature.c -177 +261

Tue Oct 12 01:05:33 EDT 2010  stevenj@alum.mit.edu
  * correct error return val for 0-dimensional integral with multiple integrands

    M ./cubature.c -3 +3

Mon Oct 11 22:03:12 EDT 2010  stevenj@alum.mit.edu
  * documentation of new feature, acknowledgements

    M ./README +37
    M ./cubature.c +8

Mon Oct 11 21:50:53 EDT 2010  stevenj@alum.mit.edu
  * whoops, correct error estimation in 1d case; now matches old cod

    M ./cubature.c -5 +15

Mon Oct 11 21:30:07 EDT 2010  stevenj@alum.mit.edu
  * bug fix in new vectorized code; now gives identical results to before

    M ./cubature.c -6 +6

Mon Oct 11 19:02:22 EDT 2010  stevenj@alum.mit.edu
  * clean up where pts/vals are allocated

    M ./cubature.c -49 +57

Mon Oct 11 18:45:36 EDT 2010  stevenj@alum.mit.edu
  * first stab at vectorized version (thanks to Dmitry Turbiner for hacking an initial prototype)

    M ./Makefile -1 +3
    M ./cubature.c -185 +225
    M ./cubature.h +13

Fri Oct  2 19:35:53 EDT 2009  stevenj@alum.mit.edu
  * use .tgz, not .tar.gz

    M ./Makefile -1 +1

Fri Oct  2 19:25:53 EDT 2009  stevenj@alum.mit.edu
  * tweak

    M ./Makefile -1 +1

Fri Oct  2 19:25:26 EDT 2009  stevenj@alum.mit.edu
  * simple Makefile

    A ./Makefile

Fri Oct  2 19:19:35 EDT 2009  stevenj@alum.mit.edu
  * document test program in README

    M ./README +43

Fri Oct  2 19:07:34 EDT 2009  stevenj@alum.mit.edu
  * added web link

    M ./README -4 +9

Fri Oct  2 19:05:01 EDT 2009  stevenj@alum.mit.edu
  * initial darcs import

    A ./COPYING
    A ./README
    A ./cubature.c
    A ./cubature.h
