7139c91d9bAdded [h4 Statistical distributions] * Student's t Perhaps switch to normal distribution as a better approximation for very large degrees of freedom?
Paul A. Bristow
2006-09-03 09:59:21 +0000
e3e83796f1Editorial and typo fixes.
John Maddock
2006-09-03 09:26:39 +0000
6c3c521a3cMore linux fixes: finally got AMD64 machines to behave correctly.
John Maddock
2006-09-02 17:56:22 +0000
53a66a80deTypo fix.
John Maddock
2006-09-02 16:16:25 +0000
84829f74e3Numerous minor patches to get tests passing on Linux x86 (32-bit) and HP-UX / itanium.
John Maddock
2006-09-02 15:05:43 +0000
0e4e5f14a6editorial work on all files and some links to Mathworld and Wiki added.
Paul A. Bristow
2006-09-02 14:02:34 +0000
4ab7fd5a36edtorial work on these qbk files - the rest still to do. Mainly spelling & punctuation.
Paul A. Bristow
2006-09-01 17:19:51 +0000
008349025eeditorial changes only except for text changed about significance & confidence level.
Paul A. Bristow
2006-09-01 16:21:12 +0000
0fe135e268after JM correction to check_df comfirmed is OK, and added output from test.
Paul A. Bristow
2006-09-01 13:20:40 +0000
567dc1bd7dUpdated error checks. Removed dead code from error-handling test.
John Maddock
2006-09-01 09:19:24 +0000
aa5fca355cadded newline at end of last line #endif // BOOST_MATH_DISTRIBUTIONS_CHI_SQUARED_HPP to avoid fatal compile error.
Paul A. Bristow
2006-08-31 17:41:19 +0000
5577b49a33Spelling and a few comments & refs.
Paul A. Bristow
2006-08-31 16:56:28 +0000
dc70a06809Added F distribution.
John Maddock
2006-08-31 16:21:17 +0000
fb79977d70Spelling corrections and re-compile & run OK and add output.
Paul A. Bristow
2006-08-31 15:19:32 +0000
3f5d624b05Comment added about big df and df = +infinity needing normal distribution.
Paul A. Bristow
2006-08-31 14:32:05 +0000
065b831b27Test update
John Maddock
2006-08-31 12:36:59 +0000
4636c32fedTest corner cases and parameter estimator.
John Maddock
2006-08-30 15:49:07 +0000
fe2ae410b9chi squared update.
John Maddock
2006-08-30 15:45:46 +0000
00a347bee0Removed remaining old style distribution code (we can get it back from cvs if necessary). Updated error handling. Added better chi-squared tests.
John Maddock
2006-08-29 16:41:07 +0000
9a3ade1320Refactored error handling a little more. Added chi-squared distribution and tests.
John Maddock
2006-08-28 12:52:08 +0000
3f52d4029bSpellings !
John Maddock
2006-08-27 09:32:00 +0000
9d7123eecfMore or less completed normal/binomial/students-t distributions. Updated docs and tests.
John Maddock
2006-08-25 18:03:48 +0000
99929b1023Fixed binomial quantile, added complements, various subtle changes to corner cases, tidied up and removed dead code. Fixed error messages in beta and gamma inverses. Updated binomial test program, refacted cdf test cases to test cdf, quantile and complements in one hit. Added accuracy tests for ibeta_inva and ibeta_invb. Added more docs.
John Maddock
2006-08-20 09:28:27 +0000
0d03447123Added overflow check.
John Maddock
2006-08-18 09:39:30 +0000
b95e6b469a1st test pdf and cdf probably OK, but others not.
Paul A. Bristow
2006-08-17 17:42:54 +0000
a23806eb611st draft - pdf and cdf probably OK, but quantile etc are not.
Paul A. Bristow
2006-08-17 17:40:55 +0000
d40fa2fb99Changed students-t quantile to be free from cancellation error. Added ibeta_inv overloads that return both x and y. Updated docs to match.
John Maddock
2006-08-16 17:33:24 +0000
a147f2aa58Fixed typo in test result message. Added ibeta inverse tests.
John Maddock
2006-08-16 12:06:10 +0000
ef43e16e47Adjusted termination conditions to cope with very large first derivatives in some regions.
John Maddock
2006-08-16 12:05:08 +0000
f965af9da3Updated incomplete beta code to calculate power terms in a more reliable way. Added explanatory comments describing what's going on.
John Maddock
2006-08-16 11:40:38 +0000
d66b96d4d6Added #include <boost/cstdint.hpp> for uintmax_t
Paul A. Bristow
2006-08-16 11:40:11 +0000
2befb07120Adjusted limits to avoid asserts.
John Maddock
2006-08-16 11:39:11 +0000
a7c3186b5aAdded fix for cdf when degrees of freedom parameter is large.
John Maddock
2006-08-16 11:38:07 +0000
4ff84d85abAdded sample programs that generate test data for incomplete beta and gamma inverses. Added asin to ntl.hpp: needed for incomplete beta inverse. Fixed limits on table size in the incomplete beta.
John Maddock
2006-08-14 17:29:59 +0000
7e902a680eAdded root-2 constant. Fixed complement code so that constructors are used now. Adjusted pre-formatted parts of tutorial. Added normal distribution plus tests.
John Maddock
2006-08-14 12:33:49 +0000
24b13c2981Comments & Quieted warnings and added test output
Paul A. Bristow
2006-08-14 11:39:28 +0000
463715c9e4quieted warmings & added test output.
Paul A. Bristow
2006-08-14 11:38:17 +0000
eacc0ceb7cquieted warnings on unreachable code..
Paul A. Bristow
2006-08-14 11:36:38 +0000
c6e4a1aabcQuieted warning unreachable code.
Paul A. Bristow
2006-08-14 11:35:20 +0000
ee24bfd3eecomments and quieted of warnings unreachable code.
Paul A. Bristow
2006-08-14 11:34:15 +0000
4e98273785Added upper cap on the number of iterations permitted in series evaluation. Fixed beta function so it's better behaved with very large arguments. Updated tests.
John Maddock
2006-08-13 17:42:17 +0000
8cc9b6d50dcommented out distributions now obselete. Need to redo all these with the distribution classes.
Paul A. Bristow
2006-08-13 11:39:58 +0000
a436522863Disable long double tests where not supported.
John Maddock
2006-08-12 18:03:54 +0000
9b1798b830Updated and proof-read docs.
John Maddock
2006-08-12 18:01:44 +0000
3ce59eaed5Fixed gcc errors (duplicate definitions of default parameters).
John Maddock
2006-08-12 18:00:54 +0000
ed7804629dUpdated to use distributions & cosmetic, & removals of warnings at level 4.
Paul A. Bristow
2006-08-12 16:09:46 +0000
9ec09012c3Added disable warnings for MSVC and output.
Paul A. Bristow
2006-08-12 15:50:29 +0000
dab4cd3c29attempted add a test for t = 0 and varying degrees of freedom - fails for big df >= 1e16.
Paul A. Bristow
2006-08-12 13:29:38 +0000
fec4b853a1*** empty log message ***
Paul A. Bristow
2006-08-12 13:28:12 +0000
50ad897d31Removed - to be replaced by version in \distributions
Paul A. Bristow
2006-08-12 11:13:59 +0000
d34cdb4311Replaced by version in \distributions.
Paul A. Bristow
2006-08-12 11:11:08 +0000
530d38dd15quieted C4610 doesn't look helpful. Added copyright JM and a refercen. Also removed a conversion warning with a static cast<RealType>, added output and commited.
Paul A. Bristow
2006-08-12 11:06:32 +0000
e70e08b902Tidied up bbv2 setup. Added images and stylesheets.
John Maddock
2006-08-11 15:31:31 +0000
38175c5aeeAdded Paul's name!!
John Maddock
2006-08-11 12:37:48 +0000
9895ce48e1Use distributions instead of dist.
Paul A. Bristow
2006-08-11 10:36:18 +0000
59a3e0b48fMoved to distributions folder.
Paul A. Bristow
2006-08-11 10:33:53 +0000
c88d67eaaaHopefully fixed graphics.
John Maddock
2006-08-11 09:33:53 +0000
fe21fb46b3More tutorial code.
John Maddock
2006-08-10 18:16:39 +0000
5c645698c3Added new style Student's t code and examples. Added tutorial for the new stats code.
John Maddock
2006-08-10 12:53:07 +0000
2dcc9167e8Updated inverse gamma tests. Updated inverse gamma code to prevent premature termination of Halley Iteration.
John Maddock
2006-08-10 09:27:19 +0000
46d9b33645Tightened up root finding tests. Added docs for roots without derivatives. Added derivatives of incomplete gamma and beta.
John Maddock
2006-08-01 17:23:10 +0000
cd90a9f0d4Updated and documented the new root finding code.
John Maddock
2006-07-31 18:05:41 +0000
eca034bb42Added tests for gamma_P_inva / gamma_Q_inva. Added docs for new inverses of gamma and beta.
John Maddock
2006-07-31 11:54:46 +0000
5e67622a7dFixed gamma_inva. Added floor and ceil termination conditions to toms748_solve. Added ibeta_inva and ibeta_invb.
John Maddock
2006-07-27 15:45:17 +0000
5e9f8e74cfAdded new equations for docs.
John Maddock
2006-07-26 13:08:02 +0000
79a8199f9cAdded new root finding algorithm toms748. Tidied up inverse code and moved files to detail directory. Updated docs and Jamfile.
John Maddock
2006-07-26 13:06:56 +0000
4ecb968fc7Begun to refactor and tighten up the tests so that results are actually regression tested as well as printed to screen. Added code for extracting the sign of a number. Fixed a few bugs and tidied up the code a little in places. Changed test_result to properly encapsulate it's contents.
John Maddock
2006-07-25 09:17:33 +0000
605742d626(merge from head) fixed typo in leading comment; removed unnamed namespace reported by inspect tool
Gennaro Prota
2006-07-24 21:49:33 +0000
7103f8c0c9fixed typo in leading comment; removed unnamed namespace reported by inspect tool
Gennaro Prota
2006-07-24 21:47:09 +0000
e1467c8da3(merge from head) same as for acos.hpp: fixed license reference text; slight changes to comments to avoid spurious inspect tool warnings
Gennaro Prota
2006-07-24 21:38:40 +0000
2875cd33b7same as for acos.hpp: fixed license reference text; slight changes to comments to avoid spurious inspect tool warnings
Gennaro Prota
2006-07-24 21:37:11 +0000
8728ec7e89(merge from head) fixed license reference text; slight changes to comments to avoid spurious inspect tool warnings
Gennaro Prota
2006-07-24 21:33:20 +0000
f31a98ad95fixed license reference text; slight changes to comments to avoid spurious inspect tool warnings
Gennaro Prota
2006-07-24 21:29:43 +0000
e12af6d280Improved tests, tightened limits, and made sure input values in the gamma tests are exact.
John Maddock
2006-07-20 12:59:17 +0000
94ebae476aUpdated generic erf version so it doesn't call the incomplete gamma directly. Refactored the incomplete gamma, improved the changeover between P and Q, added two new finite-sum expansions for integer and half integer arguments. Improved error handing in the inverse gamma.
John Maddock
2006-07-20 12:58:29 +0000
c4eceb5372Added missing license info.
John Maddock
2006-07-16 16:01:25 +0000
2760d7ad0dAdded missing license info.
John Maddock
2006-07-16 16:01:25 +0000
74b41709bfBorland C++ fix.
John Maddock
2006-07-13 09:47:18 +0000
607e6990cdAdded needed include.
John Maddock
2006-07-12 17:52:28 +0000
cc6a224a8aAdded needed #include.
John Maddock
2006-07-12 17:44:46 +0000
e8c5217846Example 1 from Wikipedia seems OK.
Paul A. Bristow
2006-07-12 16:16:34 +0000
4350382d58Test seems OK now, including inverse.
Paul A. Bristow
2006-07-12 16:14:24 +0000
162f416732Looks nearer to correct now with test and example1.
Paul A. Bristow
2006-07-12 16:13:11 +0000
7596f7bad9Update to results are corrected .hpp file error
Paul A. Bristow
2006-07-11 16:10:45 +0000
e06f1fa83bCorrected missing k+1
Paul A. Bristow
2006-07-11 16:09:04 +0000
305ca1f705comments etc PAB
Paul A. Bristow
2006-07-11 14:17:49 +0000
8417d24efdcomment lexical cast fqult added
Paul A. Bristow
2006-07-11 14:16:50 +0000
58d4299525Another unused parameter conveted to /* l */
Paul A. Bristow
2006-07-11 14:15:42 +0000
45fce41aa2Warning suppression.
John Maddock
2006-07-11 12:30:40 +0000
cda4c449a9Crush a couple more warnings.
John Maddock
2006-07-11 12:27:48 +0000
11bb475338More comments, still not tested much.
Paul A. Bristow
2006-07-10 17:41:53 +0000
86c65d8c9fAdd using declaration so that std::fabs is actually found.
John Maddock
2006-07-10 16:25:06 +0000
31fbe1eaebAdded needed math_fwd.hpp includes were needed, cleaned up code, and generally got things compiling with gcc.
John Maddock
2006-07-10 15:30:36 +0000
6307ffc6d3tested by PAB at 8 July - links and runs OK
Paul A. Bristow
2006-07-09 18:33:33 +0000
bc87a3747dAdded #pragma's where warnings couldn't be suppressed any other way. Refactored code to suppress warnings where possible.
John Maddock
2006-07-09 17:56:55 +0000
f7961cdc6dAdded missing inline specifiers.
John Maddock
2006-07-09 17:54:06 +0000
57d3e67695Initial commit
John Maddock
2006-07-09 16:40:09 +0000