Allan_Ecker
New member
- Joined
- Jul 20, 2009
- Messages
- 2
I'm trying to re-learn DiffEq from scratch to fill in holes in my knowledge, and I seem to be running into this identity that to me makes no sense. I ran into it rather early and so kind of drew a box around it and moved on, but I can't help but think it's really important to fundamental understanding:
The Sixth Edition Boyce&Diprima Elementary Differential Equations and Boundary Value Problems text solves the following first-order equation:
dy/dt + y/2 = 3/2
in a way that really confuses me. To better illustrate exactly what's confusing me, I'll use a simplified equation which has the same point of contention:
dy/dt = y
Which is equivalent to:
dy/dt * 1/y = 1
Which my textbook jumps to:
d/dt ln(y) = 1
There's some text, indicating this was done via differentiation:
dy/dt * 1/y = 1
dy/dt * (d/dy ln(y)) = 1
d/dt ln(y) = 1
Above is where I get confused. I've lived my life thus far with a blurry notion that dy and dt were standins for delta-y and delta-t, where delta is really "the limit as delta goes to zero" and if dy and dt are variables, then in the above steps dy cross-cancels with itself to yield the final answer of d/dt. But if dy is a variable, what is d/dt? Is "d" a variable? Or is d/dt an operator? If d/dt is an operator, isn't d/dy an operator too? If they're both operators and not variables, then how is this cross-multiplying step possible? Is that an identity I don't know about? Such an identity would be:
d/dt(y) * d/dy(f(y)) = d/dt(f(y))
Which then sort of slams me face-first into the fact that all the calculus identities I know are monovariate identities which don't really apply here. Help?
The Sixth Edition Boyce&Diprima Elementary Differential Equations and Boundary Value Problems text solves the following first-order equation:
dy/dt + y/2 = 3/2
in a way that really confuses me. To better illustrate exactly what's confusing me, I'll use a simplified equation which has the same point of contention:
dy/dt = y
Which is equivalent to:
dy/dt * 1/y = 1
Which my textbook jumps to:
d/dt ln(y) = 1
There's some text, indicating this was done via differentiation:
dy/dt * 1/y = 1
dy/dt * (d/dy ln(y)) = 1
d/dt ln(y) = 1
Above is where I get confused. I've lived my life thus far with a blurry notion that dy and dt were standins for delta-y and delta-t, where delta is really "the limit as delta goes to zero" and if dy and dt are variables, then in the above steps dy cross-cancels with itself to yield the final answer of d/dt. But if dy is a variable, what is d/dt? Is "d" a variable? Or is d/dt an operator? If d/dt is an operator, isn't d/dy an operator too? If they're both operators and not variables, then how is this cross-multiplying step possible? Is that an identity I don't know about? Such an identity would be:
d/dt(y) * d/dy(f(y)) = d/dt(f(y))
Which then sort of slams me face-first into the fact that all the calculus identities I know are monovariate identities which don't really apply here. Help?