Suppose that L{} is the differential operator \frac{ d }{ dx }. You could write the equation as a_k L^k{ y } = 0
So the differential equation was transformed to an algebraic one. \lambda_k are the roots of that ``polynomial''
Reverting the transformation, a solution takes the form \exp( \lambda x ) (D'Alembert)
where \exp is the exponentiation function.
The general solution is a linear combination of all the solutions. y(x) = A_k \exp( \lambda_k x )
where the A_k coefficients are determined by the initial conditions.
Notes:
_ I'd forgot the variable in the equation in the previous post.
_ I think that I'm abusing summation notation.
The idea is that repeated indexes mean to sum over them, for all their values.
By instance the inner product \sum_{k=0}^{n-1} v^2_k = v_k v_k
In this case just take it as a linear combination.
_ If \lambda is a multiple root, then there are solutions of the form x \exp( \lambda x )
_ The method can be applied to ODE linear, of constant coefficients
Yeah but we get complex roots, and so the general formula can be re-written in terms of sin and cos without imaginary numbers, and this has a specific solution to our initial values.
But as I said before this is tedious and unnecessary to anyone who knows simple harmonic motion formulae.