# Solving coupled ODE problem with explicit fft terms of the independent variables
**URL:** https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779
**Category:** New to Julia
**Tags:** ode
**Created:** [November 19, 2021, 1:47pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779 "2021-11-19T13:47:22Z")
**Posts on this page:** 8
**Page:** 1
### Author:  [@Thaegy](https://discourse.julialang.org/u/Thaegy)
#### Post date: [November 19, 2021, 1:47pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/1 "2021-11-19T13:47:22Z")
Hi all,
I have been trying to translate my MATLAB code that solves a set of coupled ODEs into Julia. I have only just started using Julia so am still quite unfamiliar with its types and the intricacies of getting some functions to behave nicely with said types. Below is the section of code I have written in Julia and below that, the code from MATLAB.
```julia
## Constants
L=10.0 # propagation distance
steps=1500.0 # number of steps to take along z
dz=L/steps # Discrete step between consecutive points along the propagation direction.
beta_2F = -1.0
beta_2S = -1.0
gamma_c = 2.0
gamma_s = 1.0
gamma_f = 1.0
s1 = 0.0
s2 = 2.0
s3 = 0.0
s4 = beta_2F/abs(beta_2S)
s5 = gamma_c/gamma_s
s6 = gamma_f/gamma_s
alpha = (test .* domega).^2.0
Omega = 0.5.*(s4.*omega .^ 2.0) .- s1.*omega
q = 250 ##
## Initial conditions
function f(x,y,z,a)
return a.*exp.(-((x.-y).^2.0)./(2*z^2.0))
end
psi_0 = sqrt(2.0*q)*sech.(sqrt(2.0*q).*(t.-30))
F_0 = f.(t,10,10,0.01).*exp.(1im*12*t)
phi_0 = fft(psi_0)
phiF_0 = fft(F_0)
ini = [phi_0 phiF_0]
z = range(0, L, length = Int64(steps + 1.0))
function ODE!(dX,X,p,t)
s2, s3, s5, s6, alpha, Omega, N = p
exp_term_psi = exp.(1im .* alpha.*t)
exp_term_phi = exp.(1im .* Omega.*t)
A = X[1:N,1] .* exp_term_psi # Do I need to specify using X[1:N,1], or can I just use X[1]?
B = X[1:N,2] .* exp_term_phi
psi = ifft(A)
phi = ifft(B)
rhs1 = 0.5 * exp(-1im * s3 * t) * s2 .* phi .* phi + psi .* abs.(psi) .* abs.(psi) + psi .* s5 .* abs.(phi) .* abs.(phi)
rhs2 = exp(1im*s3*t) * s2 .* psi .* conj(phi) + phi * s5 .* abs.(psi) .* abs.(psi) + phi .* s6 .* abs.(phi) .* abs.(phi)
rhs1 = fft(rhs1)
rhs2 = fft(rhs2)
dX[1] = 1im .* rhs1./exp_term_psi
dX[2] = 1im .* rhs2./exp_term_phi
#= dX[1] = 1im.*fft.((0.5.*s2.*(phi.^2.0).*exp.(-1im*s3.*t) .+ (abs.(psi).^2.0).*psi .+ s5.*(abs.(phi).^2.0).*psi))./exp_term_psi
dX[2] = 1im.*fft.((s2.*psi.*conj(phi).*exp.(1im*s3.*t) .+ s5.*(abs.(psi).^2.0).*phi .+ s6.*(abs.(phi).^2.0).*phi))./exp_term_phi
=#
end
p = [s2, s3, s5, s6, alpha, Omega, N];
prob = ODEProblem(ODE!,ini,(0.0,L),p);
sol = solve(prob, Tsit5());
```
The current error is:
LoadError: MethodError: Cannot `convert` an object of type Vector{ComplexF64} to an object of type ComplexF64
I’m not sure where this is an error and my debugging efforts in VSCode haven’t been very helpful at locating the problem. Am I specifying the vector variables correctly using X[1:N,1]? In MATLAB the solution is a vector of solved ODEs that I can use deval on to evaluate for any given z value I like. I’m not sure if Julia works in a similar way.
Thanks for any help in advance!
Below is the related MATLAB code, the constants are identical between the Julia and MATLAB versions:
```julia
function D = f(t,X, s2, s3, s5, s6, alpha, N, Omega)
exp_term_psi = exp(1i*alpha.*t); % For S equation
exp_term_phi = exp(1i*Omega.*t); % For F equation
psi = ifft(X(1:N).*exp_term_psi,N);
phi = ifft(X(N+1:2*N).*exp_term_phi,N);
sol = [(1i*fft(0.5*s2*(phi.^2).*exp(-1i*s3*t) + (abs(psi).^2).*psi + s5*(abs(phi).^2).*psi)./exp_term_psi),...
(1i*fft(s2*psi.*conj(phi).*exp(1i*s3.*t) + s5*(abs(psi).^2).*phi + s6*(abs(phi).^2).*phi)./exp_term_phi)];
D =[sol(:,1);sol(:,2)];
end
```
---
### Author:  [@ChrisRackauckas](https://discourse.julialang.org/u/ChrisRackauckas)
#### Post date: [November 19, 2021, 3:38pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/2 "2021-11-19T15:38:28Z")
> [@Thaegy](#):
>
> ```julia
> P_rhs1 = plan_fft.(rhs1)#; flags=FFTW.ESTIMATE, timelimit=Inf)
> P_rhs2 = plan_fft.(rhs2)#; flags=FFTW.ESTIMATE, timelimit=Inf)
>
> ```
Don’t broadcast the plan. `P_rhs1 = plan_fft(rhs1)` You want to apply it to the array, not element-wise.
---
### Author:  [@stevengj](https://discourse.julialang.org/u/stevengj)
#### Post date: [November 19, 2021, 3:59pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/3 "2021-11-19T15:59:49Z")
Note also that the whole point of `plan_fft` is to do it once and re-use it, not to recompute the plan on every timestep.
---
### Author:  [@Thaegy](https://discourse.julialang.org/u/Thaegy)
#### Post date: [November 19, 2021, 4:16pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/4 "2021-11-19T16:16:21Z")
Ah yes of course thank you, a silly mistake on my half. I have now updated my code and will edit my submission with yours and @stevengj’s suggestions.
---
### Author:  [@stevengj](https://discourse.julialang.org/u/stevengj)
#### Post date: [November 19, 2021, 4:22pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/5 "2021-11-19T16:22:24Z")
Note also that instead of `exp.(1im .* x)` it is more efficient to use `cis.(.- x)`, and simpler to do `@. cis(-x)`
---
### Author:  [@Thaegy](https://discourse.julialang.org/u/Thaegy)
#### Post date: [November 19, 2021, 4:37pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/6 "2021-11-19T16:37:58Z")
Great thank you, another useful function to know!
---
### Author:  [@baggepinnen](https://discourse.julialang.org/u/baggepinnen)
#### Post date: [November 19, 2021, 4:49pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/7 "2021-11-19T16:49:44Z")
`cis` = \cos + i \sin for your memory 😉
---
### Author:  [@Thaegy](https://discourse.julialang.org/u/Thaegy)
#### Post date: [November 19, 2021, 4:55pm UTC](https://discourse.julialang.org/t/solving-coupled-ode-problem-with-explicit-fft-terms-of-the-independent-variables/71779/8 "2021-11-19T16:55:18Z")
I was wondering where cis came from… Thanks for clarifying!