02:00 -!- Prax01D(n=Prax01D@149.99.99.135) has joined #maxima
02:23 -!- JohnFlux_(n=JohnFlux@cpc1-brig10-0-0-cust482.brig.cable.ntl.com) has joined #maxima
02:23 <JohnFlux_> Hey all!
02:23 <JohnFlux_> I'm trying to solve a second order differential equation.  I'm trying:
02:23 <JohnFlux_> (%i6) solve(m*diff(x(t),t,2) + 4*k*x(t) = m*g + 5*k*l, x(t));
02:23 <JohnFlux_> but it just returns basically what i entered
02:23 <JohnFlux_> it doesn't do the integration
02:25 <JohnFlux_> oh
02:25 <JohnFlux_> desolve seems to be what I want
02:29 <JohnFlux_> I'm reading the docs on it now
02:29 <JohnFlux_> but just want to say that I really like maxima
02:30 <JohnFlux_> Thank you,  to any developers listening ;-)
02:46 <JohnFlux_> just one thing..
02:46 <JohnFlux_> it doesn't seem to realise that  sqrt(a*b)/b    is simpler to write as sqrt(a/b)
02:58 <HaraldG> JohnFlux_: Seems maxima likes sqrt(a)/sqrt(b) most.
02:59 <HaraldG> Perhaps you are looking for radcan() ?
02:59 <JohnFlux_> HaraldG hey
02:59 <JohnFlux_> HaraldG a quick question please...    if I do   integrate(x,x);   i don't get a constant of integration
02:59 <JohnFlux_> is this on purpose?
03:00 <HaraldG> Yes, i think so.
03:01 <HaraldG> It is quite tricky to get the most general integral, so maxima just gives you special.
03:02 <HaraldG> But I think it always gives all constans when solving ODEs.
03:10 <JohnFlux_> HaraldG: btw, do you use a GUI ?
03:11 <HaraldG> Depends on what I'm doing. Usually not, but I think wxmaxima is quite nice.
03:11 <HaraldG> It has some extra features, like better latex export, too.
03:12 <JohnFlux_> ah cool
03:12 <JohnFlux_> i write up in latex, solve it myself, then checking with maxima in plain text
03:12 <JohnFlux_> i could probably use some tools to streamline that a bit
03:27 <JohnFlux_> HaraldG: can I get it to try harder to simplify an equation?
03:27 <JohnFlux_> HaraldG: one of the terms in the solution has      4 k m cos(..) /m
03:28 <JohnFlux_> i have no idea why it didn't cancel the m's
03:29 <HaraldG> Well, simple canceling should be done by the core simplifier.
03:29 <JohnFlux_> yeah it's funny
03:29 <HaraldG> Does it work, if you enter the expression yourself?
03:29 <JohnFlux_> yes
03:29 <JohnFlux_> (%i24) A*cos(sqrt(4*k/m)*t + p) + (m*g+5*k*l)/(4*k);
03:29 <JohnFlux_> (%i25) m*diff(%o24,t,2) + 4*k*%o24;
03:30 <JohnFlux_> that gives me a long equation when most of it trivally cancels out
03:31 <JohnFlux_> ooo
03:31 <JohnFlux_> it is of course totally my faul
03:31 <JohnFlux_> fault
03:32 <JohnFlux_> well, i mean it doesn't trivally cancel out
03:32 <JohnFlux_> but the problem is still there
03:33 <JohnFlux_> wait I was originally right sorry - it does cancel out
03:33 <JohnFlux_> damn this text view :P
03:33 <HaraldG> Well, I see the issue with the "m". It goes away, when I do ev(%).
03:33 <JohnFlux_> it should cancel out, rather
03:34 <HaraldG> That is, after an extra evalutation. I don't understand why this is needed ...
03:34 <JohnFlux_> although even then
03:34 <JohnFlux_> after the ev(%)   it still hasn't cancelled most of it out
03:34 <JohnFlux_> it should just give gm+5kl
03:34 <JohnFlux_> shouldn't it ?
03:35 <HaraldG> To have the simplifier handle trigonometric functions, you have to use trigsimp().
03:36 <JohnFlux_> hmm, it's not simplifying that
03:36 <HaraldG> But the core simplifier should hanlde m/m at the first run IMO.
03:36 <JohnFlux_> look at the result after ev(%)   - it has:
03:37 <JohnFlux_> 4 k ( _ + (gm+5kl)/(4k)) + 4 k _
03:37 <JohnFlux_> where _  is the cos bit
03:37 <JohnFlux_> doh
03:37 <JohnFlux_> 4 k ( _ + (gm+5kl)/(4k)) - 4 k _
03:37 <JohnFlux_> i mean
03:37 <JohnFlux_> so it should cancel the  4 k _    bits
03:38 <JohnFlux_> that's not trigometric simplification 
03:38 <HaraldG> Yes, I see your point. 
03:38 <JohnFlux_> trigsimp    does fix it though and give the right answer
03:39 <HaraldG> Hm, expand() does the right thing too.
03:39 <HaraldG> Not yet sure, if this is what I expect.
03:43 <HaraldG> Unfortunately the core simplifier is almost undocumented, i already
03:44 <HaraldG> played with it several days to get an idea what it does or does not.
03:45 <JohnFlux_> HaraldG: It's written in lisp?
03:45 <HaraldG> It seems it handles a*(c/a)-c well while it gives up on a*(b+c/a)-c-a*b.
03:45 <HaraldG> No idea if this should be considerd a bug or a feature.
03:46 <HaraldG> Yes, almost all of maxima is written in lisp.
03:46 <JohnFlux_> i think it's a bug to do   a*(b+c/a)
03:46 <JohnFlux_> that should be immediately written as  ab + c
03:46 <JohnFlux_> almost all?
03:47 <HaraldG> Some addon packages are written in maxima itself.
03:47 <JohnFlux_> ah
03:48 <HaraldG> I won't say that this is a bug that easily. 
03:48 <HaraldG> Some people would like to get rid of the core simplifier totally and handle everything with explicit function calls.
03:49 <HaraldG> The core simplifier can't be controlled (only turned off as a whole). Therefore it shouldn't do too much.
03:51 <HaraldG> I don't know it anybody has use for verbatim a*(b+c/a) ...
03:52 <HaraldG> But I have put this on my list of issues to keep in mind.
08:01 -!- Prax01D(n=Prax01D@149.99.85.87) has joined #maxima
14:03 -!- Prax01D(n=Prax01D@149.99.90.59) has joined #maxima
16:48 -!- HaraldG(n=rldprog@chello080108070152.13.11.univie.teleweb.at) has joined #maxima
19:53 -!- Prax01D(n=Prax01D@tor58-10-93-63.dynamic.rogerstelecom.net) has joined #maxima
20:09 -!- Ger895235976(n=Gert@p2p-gtn-e.ilca.ru) has joined #maxima
22:20 -!- HaraldG(n=rldprog@chello080108070152.13.11.univie.teleweb.at) has joined #maxima