PDA

View Full Version : Why no constant of integration for the integration factor?


nomadreid
Dec17-11, 01:39 AM
In applying the usual formula for y'+py=q with u=e∫p and y=(∫uq +C)/u, the results of the indefinite integral to calculate u is without a constant. Why is this? The result is not the same.
For example, suppose I take the problem
dy/dx +y = x*exp(-x). Then, in the usual manner, I get a nice exp(-x)*(x2/2 +C). But if I were to put constants in for the integration factor, I get
exp(-x)*(C1x2 +C2)
which is not the same thing.

JHamm
Dec17-11, 05:07 AM
y' + py = q
Define \displaystyle p = P'
u = e^{\int p\ dx}
= Pe^C \Rightarrow u = AP
Then \displaystyle (yu)' = uq
y = \frac{\int uq\ dx}{u}
= \frac{\int APq\ dx}{AP}
= \frac{A}{A}\frac{\int Pq\ dx}{P}
So adding the constant doesn't change anything :)

nomadreid
Dec17-11, 06:06 AM
Thanks, JHamm.