~peterpall/maxima/+git/maxima.code:rtoy-compute-subdirs-at-runtime

Branch merges

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related charm recipes

08cdc78... by Raymond Toy <email address hidden> on 2023-07-16

Merge branch 'master' into rtoy-compute-subdirs-at-runtime

e3b7277... by Barton Willis <email address hidden> on 2023-07-15

o Make change that is suggested in source code about calling limit at the top of limit2.

Tested with SBCL 2.2.2; the testsuite & share testsuite pass.

53f6907... by Raymond Toy <email address hidden> on 2023-07-14

Merge branch 'master' into rtoy-compute-subdirs-at-runtime

24a57b6... by Barton Willis <email address hidden> on 2023-07-14

Fixes for rtest_limit_extra bugs 52, 64, 67, 234, 277, 278, and 279.

Specifically, fixes for the reported bugs

  #3846 limit gives quotient by zero error
  #4147 integrate(log(sin(x))/cos(x),x,0,%pi/2)
  #4138 limit((1+1/x)^x,x,6/5) is "und" in complex domain
  #4133 limit(abs(cos(a*x)),x,inf)
  #4103 limit(acot(x),x,0) should be IND (not UND)
  #3932 inconsistent limit results with trig

o Blend functionality of factosimp into extra-simp.

o Blend functionality of limitsimp into extra-simp. But we no longer do
  the X^Y --> exp(X^Y) transformation when both X & Y depend on the limit
  variable. The limit code doesn't need this transformation and it caused
  a bug with a two sided limit.

o Blend functionality of tansc into extra-simp. The limit code and the definite
integration code call tansc.

o New version of function /#alike. This version is more efficient.

o Improvements to stirling0.

o At one time when simplimplus1 failed to find a limit, this function
applied ratsimp and tried again. This scheme didn't fix any testsuite
bugs and it slowed the code, so I removed it.

o Function simplimplus1 now gathers zerob & zeroa terms and it
  attempts to condense multiple infinity terms into a single term. Finally
  there is new code for attempting to simplify multiple ind terms into a
  single term.

o New simplifications for extra-simp:

  (1) sin(x)^2+cos(x)^2 --> 1 simplification
  (2) log(negative number) --> log(-negative number) + %i*%pi
  (3) apply cos(X)^2 + sin(X)^2 --> 1, when X depends on var
  (4) convert, gamma functions, binomial coefficients, and beta functions to
       factorial form
  (5) do some reciprocal function transformations; for example csc(X) --> 1/sin(X)
  (6) do transformations similar to acsc(X) --> asin(1/X).
  (7) convert Fibonacci functions to power form.

  Tested with SBCL 2.2.2 & Clozure Common Lisp Version 1.12 (v1.12) WindowsX8664.
  The testsuite & share testsuite pass.

f2b6cd0... by David Billinghurst <email address hidden> on 2023-07-14

Quote '%generalized_lambert_w in definition of $generalized_lambert_w.

e762c5e... by David Billinghurst <email address hidden> on 2023-07-13

Correct typo in reference s/Verebic/Veberič/

be99a13... by David Billinghurst <email address hidden> on 2023-07-13

Replace simplify with ftake* in $lambert_w and $generalized_lambert_w.

4caa61f... by Raymond Toy <email address hidden> on 2023-07-12

Replace (simplify (list '(%beta)) ...) with (ftake* '%beta ...)

It's nicer to use ftake* to replace the calls to simplify. It's less
typing for one, but it also makes it clearer what we want and more
consistent through the code base so that if we need to do something
with ftake*, it's automatically handled everywhere.

af220c4... by Raymond Toy <email address hidden> on 2023-07-12

Remove old commented out beta simplifer that was replaced.

446bc61... by Raymond Toy <email address hidden> on 2023-07-12

Replace simpbeta with def-simplifier for beta

This means we need to replace $beta with %beta everywhere.