- Jun 25, 2010
-
-
Liam Healy authored
-
- Jun 10, 2010
-
-
Liam M. Healy authored
The expansion of a foreign structure-by-value function did not substitute the foreign function name. It turns out this didn't matter because cffi::parse-name-and-options used by fsbv:defcfun simply ignored the symbol 'GSL-NAME for the foreign function name and pulled the actual name out of the symbol-name for the Lisp symbol of the function name. This fix makes it clearer in the macroexpansion however.
-
- Jun 08, 2010
-
-
Liam M. Healy authored
Dependencies line for clbuild is now fixed; tests count and source.
-
- Jun 06, 2010
-
-
Liam Healy authored
The :cl-array argument to make-marray is obsolete and has been eliminated; the :data argument has replaced it (the argument must have been make by c-array:make-ffa). Added initial-element and initial-contents arguments to grid:make-grid-data method.
-
- Jun 02, 2010
-
-
Liam M. Healy authored
-
Liam M. Healy authored
The gamma tests that were failing under Debian Lenny/SBCL 1.0.18/GSL 1.11 pass under Debian squeeze/SBCL 1.0.36.25/GSL 1.14, so they are now uncommented. The exponential-function test that failed in the first case passes in the second. TOTAL: 4062 assertions passed, 2 failed, 0 execution errors.
-
Liam Healy authored
The macro #'assert-to-tolerance was not correct but now maps to sf-check-results. The order of arguments in the sf-frac-diff call in sf-check-single has reversed to match test_sf_check_result in the GSL code. The remaining exponential-function tests have been ported except for the one involving gsl_sf_exprel_n_CF_e which hasn't been ported. All the tests except one pass.
-
- Jun 01, 2010
-
-
Liam M. Healy authored
Fixed the argument order to an exponential-functions test. Error estimates used in special function calculations are optional in sf-check-results. There is an exponential-functions test which fails even though it is identical to a GSL test; commented out.
-
Liam M. Healy authored
-
Liam Healy authored
The return value from #'exp-err-scaled should be sf-result-e10, now fixed. Test macros renamed, made to match their corresponding GSL versions better, and moved to augment.lisp. Some of the GSL exponential function tests are now ported, and two fail.
-
- May 30, 2010
-
-
Liam Healy authored
-
Liam Healy authored
The special function tests now match much better the GSL (C) tests, and the gamma function tests ported from C all pass except for a few. These few also fail in small C program tests, but for some reason pass in the GSL "make check" test.
-
- May 28, 2010
-
-
Liam Healy authored
The special function test for gamma from GSL are in place, and 32 fail, some by a substantial amount. Some are commented out.
-
- May 27, 2010
-
-
Liam M. Healy authored
Modified the parser converter for semi-automatic conversion of GSL tests for gamma special function. Still need some work to handle multiple values.
-
Liam Healy authored
-
- May 26, 2010
-
-
Liam M. Healy authored
The cl-ppcre C float parser is matching to nothing because there's no single part of a float that is required; d. .d and d all possible, so none of digit-before-decimal, decimal, or digit-after-decimal are required. Workaround is to check for length of 'mant and just return the string if it's zero.
-
Liam M. Healy authored
-
Liam Healy authored
Split convert-gsl-test into multiple function calls for modularity. Improve number parsing, still a work in progress. Arguments may be selected.
-
- May 25, 2010
-
-
Liam Healy authored
Rewrote testpdf so that it takes a function of the random variable, and the list of distribution name and keyword-argument pair. Added new tests levy and exponential-power; several more (beta, chi-squared, exponential) written but commented out because they take an unreasonably long time to complete; these require further investigation.
-
- May 24, 2010
-
-
Liam Healy authored
Convert the tests from cdf/test_auto.c for random distribution; this now gives a total number of tests more than 3600: TOTAL: 3644 assertions passed, 5 failed, 2 execution errors.
-
- May 23, 2010
-
-
Liam Healy authored
-
- May 22, 2010
-
-
Liam Healy authored
Added chi-squared (which was not loaded at all), exponential, exponential-power, fdist, gamma-randist, tdist. Introduced new macros assert-posinf and assert-neginf to check for expected infinite results. Improved the automatic test converter and noted its known failings.
-
Liam Healy authored
The tests for #'ugaussian-p, #'ugaussian-q, #'ugaussian-pinv, #'ugaussian-qinv are automatically converted from the GSL tests instead of converted from the examples.
-
Liam Healy authored
Convert the cdf/test.c GSL test file to a list of CL forms. This requires cl-ppcre and iterate, and is not normally loaded with either gsll or gsll-tests.
-
Liam Healy authored
Introduced a new macro #'assert-to-tolerance to be used in lisp-unit:define-test forms that takes three arguments: form expected-value tolerance. This replaces test-tol6 that was defined in tdist.lisp. Defined the tolerance constants +test-tol6+ the exact same way as GSL, but grovelling machine.h where GSL_DBL_EPSILON is defined.
-
- Apr 26, 2010
-
-
Liam M. Healy authored
-
Liam Healy authored
In the nonlinear least squares solver, make 'covariance an optional argument to #'ls-covariance. Make the first argument the solver object instead of a pointer, so the jacobian is computed before being passed to the foreign function.
-
Russell Kliese authored
Fixed nonlinear least squares fitting to allow its use outside the gsl package * Updated fit-test-delta function to take take the solver object in order to avoid passing it raw pointers * Exported the jacobian function
-
Liam Healy authored
-
- Apr 20, 2010
-
-
Liam M. Healy authored
Improve the message from obsolete-gsl-version, telling the user to remove fasls and recompile after a GSL upgrade.
-
- Apr 18, 2010
-
-
Liam Healy authored
Use GSL tests for random distributions. Port all tdist tests; some of laplace.
-
- Apr 17, 2010
-
-
Liam Healy authored
-
- Apr 15, 2010
-
-
Liam Healy authored
-
- Apr 14, 2010
-
-
Liam M. Healy authored
The GSL function gsl_sf_elljac_e writes sn, cn, dn values to doubles for which the pointer has been passed in, as shown in /usr/include/gsl/gsl_sf_elljac.h, and not to sf-result as one might be led to conclude from the "_e" at the end of the name. This is now fixed in GSLL, so that the tests, which sometimes passed and sometimes didn't depending on what was run before it, now should always work. The single non-error test been replaced by the 35 GSL tests in gsl/specfunc/test_sf.c, and they all pass in SBCL Linux AMD64.
-
- Apr 10, 2010
-
-
Liam Healy authored
Jame Wright pointed out that the function trivial-example-step tried to apply the function #'sample to what turns out to be a foreign pointer, but it should take a CL object. This is because we are making a random number generator, passing the foreign pointer to GSL, and then it is passing that pointer back to our routines so that we can use it in the step function. This has been changed so that there is a special variable cl-generator which is bound to the CL object in #'simulated-annealing, and then used in the step function, while the foreign pointer being returned to us is ignored.
-
- Apr 05, 2010
-
-
James Wright authored
-
- Mar 28, 2010
-
-
Liam Healy authored
Extract c-pointer of interpolation arrays xa, ya before passing, instead of trying to pass the CL object directly to GSL. Fixes error http://common-lisp.net/pipermail/gsll-devel/2010q1/000437.html reported by Mirko Vukovic.
-
- Mar 14, 2010
-
-
Liam Healy authored
Creating marrays with grid:make-grid would fail because grid:make-grid-data incorrectly passed a NIL if :initial-element was not specified.
-
Liam Healy authored
Upcase string before interning in #'data-class-name, so that #m will work correctly even if *print-case* is set to :downcase. Bug reported by Leo.
-
- Feb 07, 2010
-
-
Liam Healy authored
-