1. 01 Jan, 2017 1 commit
    • New argument to defmobject · dd2e7673
      Add key argument :initialize-when-making to defmobject, which controls
      whether initialization (by calling reinitialize-instance or some other
      function) happens when the make-* function defined by this macro is
      called. Choices are
        T=yes always
        NIL=no
        :default-T=optional (default T)
        :default-NIL=optional (default NIL)
      and the default is T to be consistent with previous behavior.
      
      This is used in the newly defined defmobject 'permutation.
      Liam M. Healy committed
  2. 20 Nov, 2016 1 commit
    • Load GSL lib v19 in preference to v0 · cb9c6bfd
      Load GSL library version 19 (libgsl.so.19) in preference to version 0,
      if it is available. This version, corresponding to GSL 2.0+, loads but
      functions whose API changed have not been tested. This required a fix to
      have-at-least-gsl-version which was incorrectly comparing version
      numbers.
      Liam M. Healy committed
  3. 07 Aug, 2016 1 commit
  4. 06 Aug, 2016 1 commit
    • Start API changes for GSL 2.0 · 2d3c6598
      Release 2.0 of the GSL library changed the API for many functions, and
      this means that the previous versions of GSLL will not load for this
      release of the library. This is a start at handling these changes, with
      an initial focus on permitting GSLL to load and not yet insuring that
      the changed defintions function correctly.
      
      There are two such updates made here:
      1) The elimination of multiple arguments in gsl_multifit_linear_svd, and
      consequently an internal change to #'linear-mfit-svd.
      2) The elimination of the argument n in gsl_sf_ellint_D_e and
      consequently in #'elliptic-integral-D.
      
      This report is helpful in identifying the GSL API changes:
      http://abi-laboratory.pro/tracker/compat_report/gsl/1.16/2.0/75ebf/abi_compat_report.html
      Liam M. Healy committed
  5. 15 Jun, 2016 1 commit
  6. 12 Jun, 2016 1 commit
    • Add Monte Carlo parameter setting · bcde4754
      With GSL 1.13, the GSL API defines a way to set the parameters for the
      MISER and VEGAS Monte Carlo methods. This commit defines two generic
      functions, #'parameter and #'(setf parameter), to get and set the
      parameters, respectively.  It loads but has not been tested.
      
      The previous way of setting the parameters, with macros
      > #'miser-parameter and #'vegas-parameter, are conditional to
      > #+obsolete-gsl and so do not normally load.
      Liam M. Healy committed
  7. 06 Dec, 2015 1 commit
  8. 24 Aug, 2015 1 commit
  9. 02 Aug, 2015 1 commit
  10. 02 Jun, 2015 1 commit
  11. 23 Jan, 2015 1 commit
  12. 22 Jan, 2015 1 commit
  13. 16 Feb, 2014 1 commit
  14. 25 Nov, 2013 1 commit
  15. 18 Nov, 2012 1 commit
    • Substitute for dim0 and dim1 in funcallables when scalarsp=NIL · 127d3d02
      Previously, the funcallable form made by #'make-funcallable-form (for
      e.g. ODE solvers) when scalarsp=T involved substituting the symbols
      'dim0 and 'dim1, but when scalarsp=NIL, would not do the substitution.
      This caused an error when those unevaluated and unbound symbols
      appeared in the lambda.  By altering #'faify-form with an additional
      'dimension-values argument and calling 
      (value-from-dimensions argspec dimension-values) 
      instead of 
      (parse-callback-argspec argspec 'dimensions), 
      this problem is fixed.
      Liam M. Healy committed
  16. 19 Feb, 2012 1 commit
  17. 13 Jan, 2012 1 commit
  18. 30 Oct, 2011 1 commit
  19. 29 Oct, 2011 1 commit
    • Fix sf-result* definitions, eliminate ffexpand, fsbv:object · 121e475d
      Fix sf-result and sf-result-e10 translation methods, and use in
      complex-with-error and values-with-errors.  Some usage in gamma.lisp
      updated, but there will be a need to replace sf-result with (:struct
      sf-result) in all declarations for many special functions.
      
      I removed ffexpand and everything it depended on because there is no
      reason to make a side defcfun for call-by-value functions, and if FSBV
      is missing CFFI already emits an error.  All usage of fsbv:object and
      any thing from the FSBV package is now gone.
      Liam M. Healy committed
  20. 24 Oct, 2011 1 commit
    • Eliminate fsbv:object for built-in types · 639a5f35
      Eliminate fsbv:object for built-in types; started conversion of
      special function result (sf-result*) structures but this requires a
      modification to CFFI.  Simulated annealing structure called by value.
      Use cffi-fsbv's sizet definition; this imposes a dependence on that
      system.  No compilation attempted yet.
      Liam M. Healy committed
  21. 18 Oct, 2011 1 commit
  22. 20 Aug, 2011 1 commit
  23. 30 Jan, 2011 1 commit
  24. 13 Jan, 2011 1 commit
  25. 12 Jan, 2011 2 commits
  26. 10 Jan, 2011 2 commits
  27. 01 Jan, 2011 1 commit
  28. 19 Dec, 2010 3 commits
  29. 16 Dec, 2010 1 commit
  30. 12 Dec, 2010 1 commit
  31. 04 Dec, 2010 2 commits
  32. 28 Nov, 2010 1 commit
  33. 27 Nov, 2010 2 commits
  34. 25 Nov, 2010 1 commit
    • New macros #'maref and #'(setf maref) · 9becff94
      New macros #'maref and #'(setf maref) get and set elements of the GSL
      array directly from the mpointer.  These are not used by users, but in
      the make-funcallable-form expansion of callbacks when scalarsp = T is
      specified.  This fixes an error in the argument order in the (setf
      get-value) generic function that this replaces.
      Liam Healy committed