MATLAB.jl is a great tool for interoperability between MATLAB and Julia, but it has one problem: the C MATLAB Engine API it uses doesn’t expose the ability to connect to open MATLAB sessions the way that the Python and C++ APIs do. Since I constantly found myself wanting to transfer data between my Julia and MATLAB sessions, I decided to create SharedMATLABEngine.jl. It uses PyCall to access the Python MATLAB Engine API matlab.engine.
The intention is not to replace MATLAB.jl, as that is a much more complete project than I’m intending for this to be. It’s really meant just to give you one thing: an embedded and shared MATLAB command prompt in your Julia REPL.
Here’s a quick overview of the functionality from the README:
Overview
To begin using SharedMATLABEngine, import the library and call the function connect_matlab(engine_name). To see a list of available MATLAB sessions, call find_matlab(). Alternitavely, you can call matlab.engine.engineName in MATLAB to see the name of that specific session.
julia> using SharedMATLABEngine
julia> eng = connect_matlab("MATLAB_25596"); # Get from matlab.engine.engineName in MATLAB
REPL mode SharedMATLABEngine initialized. Press > to enter and backspace to exit.
Now the MATLAB command line can be accessed from the Julia REPL by typing >.
>> a = magic(3)
a =
8 1 6
3 5 7
4 9 2
Julia variables can be interpolated into MATLAB commands via the $ operator.
>> $a21 = a(2,1)
3.0
>> b = {zeros($a21), 'some words'}
b =
1×2 cell array
{3×3 double} {'some words'}
>> [$z, c] = b{:}
c =
'some words'
julia> (a21, z)
(3.0, [0.0 0.0 0.0; 0.0 0.0 0.0; 0.0 0.0 0.0])
MATLAB command outputs can be accessed in Julia through the mat" string macro or the Engine instance’s workspace field.
julia> b = a21 .+ mat"a"
3×3 Array{Float64,2}:
11.0 4.0 9.0
6.0 8.0 10.0
7.0 12.0 5.0
julia> sqrt.(eng.workspace.a)
3×3 Array{Float64,2}:
2.82843 1.0 2.44949
1.73205 2.23607 2.64575
2.0 3.0 1.41421