[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
## [Axiom-developer] Infinite loop computing Laplace transform

**From**: |
Waldek Hebisch |

**Subject**: |
[Axiom-developer] Infinite loop computing Laplace transform |

**Date**: |
Mon, 15 Jan 2007 02:46:49 +0100 (CET) |

I have looked at bug 101. At the end of LaplaceTransform.locallaplace
we have:
-- last chance option: try to use the fact that
-- laplace(f(t),t,s) = s laplace(g(t),t,s) - g(0) where dg/dt = f(t)
elem?(int := lfintegrate(f, t)) and (rint := retractIfCan int) case F =>
fint := rint :: F
-- to avoid infinite loops, we don't call laplace recursively
-- if the integral has no new logs and f is an algebraic function
empty?(logpart int) and algebraic?(f, t) => oplap(fint, tt, ss)
ss * locallaplace(fint, t, tt, s, ss) - eval(fint, tt = 0)
The recursion here is responsible for infinite loop when computing
laplace(log(z), z, w). Namely, at this point f = log(z). Integrating
we get z*log(z) - z. First part of locallaplace uses linearity
and recurses with f = z*log(z). Again first part factors out z and
recurses with f = log(z)...
ATM I am unable to find a function such that this fragment helps. OTOH
the fragment is quite likely to cause infinite recursion so I consider
disabling it.
--
Waldek Hebisch
address@hidden

[Prev in Thread] |
**Current Thread** |
[Next in Thread] |

**[Axiom-developer] Infinite loop computing Laplace transform**,
*Waldek Hebisch* **<=**