Re: Re: Re: Bug?

*To*: mathgroup at smc.vnet.net*Subject*: [mg16062] Re: [mg15899] Re: [mg15854] Re: Bug?*From*: "Atul Sharma" <mdsa at musica.mcgill.ca>*Date*: Tue, 23 Feb 1999 03:45:19 -0500*Organization*: McGill University Computing Centre*References*: <7allmr$gtg$1@dragonfly.wolfram.com>*Sender*: owner-wri-mathgroup at wolfram.com

For what it's worth, I get the same error with Mathematica 3.01.1 running under Win95. Strangely, if I specify the base E in the Log function, I don't get this error message, but an incorrect answer. Integrate[E^x/Log[x, E], x] Out[12]= -ExpIntegralEi[x] + E^x Log[x] Differentiating this with respect to x yields E^x Log[x], which the wrong sign (the correct answer appears if you use Integrate[E^x/Log[x,-E],x]!!!) This was an unmodified 'fresh' kernel session, so I'm also perplexed as to why this is occuring. A. Sharma Dennis Swaney wrote in message <7allmr$gtg$1 at dragonfly.wolfram.com>... >BobHanlon at aol.com wrote: >> >>In a message dated 2/12/99 8:49:55 PM, ck at kagi.com writes: >> >>>> This does not occur in any Mathematica I have. Try it again. >>>> The error message looks like you may have mistyped. >>>> >>>> Moderator >>>> >>>> > I have found something strange in Mathematica: >>>> > >>>> > Integrate[Exp[x]/Log[x],x] >>>> > >>>> > which yields the following error message: >>>> > >>>> > Log::"argt": >>>> > "\!\(Log\) called with \!\(0\) arguments; \!\(1\) or \!\(2\) >>>> > arguments \ >>>> > are expected." >>>> >