I have updated to the latest version of package NeuralPDE and used the res = Optimization.solve(prob,OptimizationOptimJL.BFGS(); callback = callback, maxiters = 1500) to solve the ODE system on Jupyter notebook. I got the message:
MethodError: no method matching cos(::Matrix{Float64})
You may have intended to import Base.cos
This issue did not occur when I used the older version (I forgot the exact version that I used before the update).
How could I fix this error?
Does this happen after a fresh start? It is possible you have re-defined cos somewhere that would give this error. For example, this works:
C:\Users\User>julia
_
_ _ _(_)_ | Documentation: https://docs.julialang.org
(_) | (_) (_) |
_ _ _| |_ __ _ | Type "?" for help, "]?" for Pkg help.
| | | | | | |/ _` | |
| | |_| | | | (_| | | Version 1.9.2 (2023-07-05)
_/ |\__'_|_|_|\__'_| | Official https://julialang.org/ release
|__/ |
julia> A=rand(5,5)
5×5 Matrix{Float64}:
0.909718 0.108619 0.995858 0.234943 0.0742485
0.965059 0.638944 0.0691136 0.618132 0.436923
0.74908 0.351674 0.0514817 0.66934 0.992477
0.211503 0.774835 0.134928 0.632693 0.0331376
0.226476 0.294484 0.88037 0.554454 0.154767
julia> cos(A)
5×5 Matrix{Float64}:
0.427955 -0.166231 -0.331578 -0.324861 -0.400794
-0.560274 0.641195 -0.496704 -0.388002 -0.101208
-0.394016 -0.35062 0.376286 -0.438875 -0.0776861
-0.367901 -0.389009 -0.0493614 0.654608 -0.148705
-0.385263 -0.322394 -0.0887973 -0.417086 0.608369
but if you first defined some new method for cos, it fails:
C:\Users\User>julia
_
_ _ _(_)_ | Documentation: https://docs.julialang.org
(_) | (_) (_) |
_ _ _| |_ __ _ | Type "?" for help, "]?" for Pkg help.
| | | | | | |/ _` | |
| | |_| | | | (_| | | Version 1.9.2 (2023-07-05)
_/ |\__'_|_|_|\__'_| | Official https://julialang.org/ release
|__/ |
julia> cos(A::String) = A # just any definition
cos (generic function with 1 method)
julia> cos(rand(5, 5))
ERROR: MethodError: no method matching cos(::Matrix{Float64})
You may have intended to import Base.cos
Closest candidates are:
cos(::String)
@ Main REPL[1]:1
Stacktrace:
[1] top-level scope
@ REPL[2]:1
(P.S. cos(A) is compute the matrix cosine of A rather than cos of each element of A, which is cos.(A), just in case that’s what you intend to do.)
But before I update all packages and Julia version 1.9.2, this error did not occur. Actually, I use the cos and sin function for variables in the ODE problem and use NeuralPDE to solve it.
In the model, if I remove the equation with the sin and cos functions it can run well.
the equations with sin and cos are:
D(omega1(t)) ~ (TM1(t)-(E1^2*Y1[1,1].re + E1*E2*sin(delta1(t)-delta2(t))*Y1[1,2].im + E1*E2*cos(delta1(t)-delta2(t))*Y1[1,2].re + E1*E3*sin(delta1(t)-delta3(t))*Y1[1,3].im +
E1*E3*cos(delta1(t)-delta3(t))*Y1[1,3].re))/(2*data["H"][1])*omega_s,
#
D(omega2(t)) ~ (TM2(t)-(E2^2*Y1[2,2].re + E1*E2*sin(delta2(t)-delta1(t))*Y1[2,1].im + E1*E2*cos(delta2(t)-delta1(t))*Y1[2,1].re + E2*E3*sin(delta2(t)-delta3(t))*Y1[2,3].im +
E2*E3*cos(delta2(t)-delta3(t))*Y1[2,3].re))/(2*data["H"][2])*omega_s,
#
D(omega3(t)) ~ (TM3(t)-(E3^2*Y1[3,3].re + E3*E1*sin(delta3(t)-delta1(t))*Y1[3,1].im + E3*E1*cos(delta3(t)-delta1(t))*Y1[3,1].re + E3*E2*sin(delta3(t)-delta2(t))*Y1[3,2].im +
E3*E2*cos(delta3(t)-delta2(t))*Y1[3,2].re))/(2*data["H"][3])*omega_s.
how could I fix this error?
yes and got the same message:
D(omega1(t)) ~ (TM1(t) - (E1^2*Y1[1,1].re + E1*E2*sin.(delta1(t)-delta2(t))*Y1[1,2].im + E1*E2*cos.(delta1(t)-delta2(t))*Y1[1,2].re + E1*E3*sin.(delta1(t)-delta3(t))*Y1[1,3].im +
E1*E3*cos.(delta1(t)-delta3(t))*Y1[1,3].re))/(2*data["H"][1])*omega_s,
#
D(omega2(t)) ~ (TM2(t) - (E2^2*Y1[2,2].re + E1*E2*sin.(delta2(t)-delta1(t))*Y1[2,1].im + E1*E2*cos.(delta2(t)-delta1(t))*Y1[2,1].re + E2*E3*sin.(delta2(t)-delta3(t))*Y1[2,3].im +
E2*E3*cos.(delta2(t)-delta3(t))*Y1[2,3].re))/(2*data["H"][2])*omega_s,
#
D(omega3(t)) ~ (TM3(t) - (E3^2*Y1[3,3].re + E3*E1*sin.(delta3(t)-delta1(t))*Y1[3,1].im + E3*E1*cos.(delta3(t)-delta1(t))*Y1[3,1].re + E3*E2*sin.(delta3(t)-delta2(t))*Y1[3,2].im +
E3*E2*cos.(delta3(t)-delta2(t))*Y1[3,2].re))/(2*data["H"][3])*omega_s];
MethodError: no method matching cos(::Matrix{Float64})
You may have intended to import Base.cos
Closest candidates are:
cos(::ForwardDiff.Dual{T}) where T
@ ForwardDiff C:\Users\htran\.julia\packages\ForwardDiff\vXysl\src\dual.jl:238
cos(::DualNumbers.Dual)
@ DualNumbers C:\Users\htran\.julia\packages\DualNumbers\5knFX\src\dual.jl:327
cos(::Float64)
@ NaNMath C:\Users\htran\.julia\packages\NaNMath\ceWIc\src\NaNMath.jl:9
...MethodError: no method matching cos(::Matrix{Float64})
You may have intended to import Base.cos
Closest candidates are:
cos(::ForwardDiff.Dual{T}) where T
@ ForwardDiff C:\Users\htran\.julia\packages\ForwardDiff\vXysl\src\dual.jl:238
cos(::DualNumbers.Dual)
@ DualNumbers C:\Users\htran\.julia\packages\DualNumbers\5knFX\src\dual.jl:327
cos(::Float64)
@ NaNMath C:\Users\htran\.julia\packages\NaNMath\ceWIc\src\NaNMath.jl:9
...
Yes, this should be automatically broadcasting, the __dot__ function that does this seems really picky about when it wants to work properly however.
I’ll add a test for this and check on 1.9.2
Was this error resolved? I get this error while running the neuralPDE example from documentation. Introduction to NeuralPDE for PDEs · NeuralPDE.jl
It was solved a few months back. What versions are you using?