MathGroup Archive 2003

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Total Derivative

  • To: mathgroup at smc.vnet.net
  • Subject: [mg42097] Re: Total Derivative
  • From: "Will Self" <wself at msubillings.edu>
  • Date: Thu, 19 Jun 2003 03:59:02 -0400 (EDT)
  • References: <bcp156$7tu$1@smc.vnet.net>
  • Sender: owner-wri-mathgroup at wolfram.com

"Joel storch" <jstorch at earthlink.net> wrote in message
news:bcp156$7tu$1 at smc.vnet.net...
> Dt[x y,x,y] gives 1+dx/dy dy/dx
>
> I would expect: 1+dx/dy dy/dx + x d/dy (dy/dx)
>
> Can someone explain this result ?

Dt[f, x, y] is supposed to give d/dy(df/dx).  Check the Help on Dt to verify
this.  Where I write d, you should read curly-d, the partial derivative.

Then Dt[x y, x, y] should mean d/dy(d(x y)/dx)

Start with the inside, d(x y)/dx.  By the product rule, this is
x dy/dx + y

Now do the derivative of this answer with respect to y

d/dy (x dy/dx + y) = x d^2y/(dydx) + dy/dx dx/dy  + 1
                    |---------------------------|
   product rule -->   1st  D 2nd      2nd  D 1st

But because mixed partials are equal, d^2y/(dydx) = d^2y/(dxdy) =
d/dx(dy/dy) = d/dx(1) = 0.

So all you're left with is dy/dx dx/dy + 1



  • Prev by Date: Re: Diophant equation
  • Next by Date: Find a dood function for fitting data
  • Previous by thread: Re: Total Derivative
  • Next by thread: Exporting animation into power point