Solving an ODE containing a function of the independent variable only known at discrete points
Solving an ODE containing a function of the independent variable only known at discrete points
I'm a little new at Mathematica and I'm trying to solve the following ordinary differential equation:
$qquad xfracdy(x)dx+y=-p(x)$.
I have data (in a list) for $x$, and the corresponding values of $p(x)$ (also in a list). I need to solve the ODE to get a list $y(x)$ for the $x$s in the 1st list. I've been trying to use NDSolve, like so:
NDSolve
NDSolve[x*y'[x] + y[x] == -p, y, x, a, b]
with no luck.
Is this possible in Mathematica? Is there a way to do this?
@Lotus yes I guess so!
– zack
Sep 5 '18 at 3:40
Then here is what you do: Create a cost function with your differential equation (NDSolve etc..) and the Norm between the solution and your data. Use NMinimize to minimize the cost function to get y(x) which should be a good fit
– Lotus
Sep 5 '18 at 3:42
@Lotus This is a misunderstanding. If I am not mistaken, OP has just some discrete data instead a function for the excitations
p (the right hand side) and wants to simply solve the ODE for y.– Henrik Schumacher
Sep 5 '18 at 6:20
p
y
2 Answers
2
Why not try to fit p(x) to the data and then using DSolve? Since you did not make a MWE, I made up some data
p(x)
DSolve
ClearAll[x, y, p]
xData = 1, 2, 3, 4;
pData = 1, 4, 9, 16;
data = Transpose[xData, pData];
p[x_] = Fit[data, 1, x, x^2, x]; (*change fit as needed*)
DSolve[x y'[x] + y[x] == -p[x], y[x], x]

Should the data be free of noise, using
Interpolation instead of Fit also comes to mind.– Henrik Schumacher
Sep 5 '18 at 5:30
Interpolation
Fit
It is a very simple equation and I recommend to first solve it analytically. It can be done by the, say, the so-called, u*v method. The solution is as follows:

Now one can directly integrate it numerically by summing up the areas of the trapezoids formed by the points x1,x2,p1 and p2. Alternatively, one can use the advice of @Nasser and integrate the fitted function, which is easier.
Have fun!
Thanks for contributing an answer to Mathematica Stack Exchange!
But avoid …
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
Some of your past answers have not been well-received, and you're in danger of being blocked from answering.
Please pay close attention to the following guidance:
But avoid …
To learn more, see our tips on writing great answers.
Required, but never shown
Required, but never shown
By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.
In other words are you trying to fit the differential equation to your data ?
– Lotus
Sep 5 '18 at 3:33