psrkwrap: A Jupyter Notebook repository from SalvadorBrandolin

Julia 1.8.3 and Python 3.10

Julia: Clapeyron v0.5.7 IJulia v1.23.3 PackageCompiler v2.1.5 PyCall v1.95.1 PyPlot v2.11.0

Python julia==0.6.0 numpy==1.21.1 scipy==1.10.0

The pyjulia library is used to make calls to Julia and evaluate the PRSK eos functions easily and quickly, without dealing directly with Julia and memorizing all the functions. Note that using the Python wrapper introduces an overhead that can make the execution up to 10 or 100 times slower.

The idea of this mini-library is to perform the flashes for my PhD with ease and convenience.

As of December 27, 2022, Clapeyron does not have the following functions implemented:

Ideal gas heat capacity DIPPR equation 107 Mathias Coppeman's ${\alpha}$ function. Defining DIPPR107 is complicated because it requires defining the ideal Helmholtz free energy function. Thus, it is easier to take the DIPPR function and adjust the ReidIdeal cubic polynomial and use that as the cp function.

On the other hand, for Mathias Coppeman's function, there is no other option.

This function is implemented in the mathiascopeman.jl file, imitating the other alpha functions in Clapeyron. It has been tested.

The installation is not very complex but has its quirks.

INSTALLATION

Easy way
In your python interpreter:

from psrk import install

install()

Manual way:

Install Julia. This is a bit cumbersome; you need to download the Julia folder and add it to the .bashrc path. For example: export PATH="$PATH:~/julia-1.8.3/bin" or the desired path for Julia...
From the terminal, run:

    julia

Julia should run if it is installed correctly. Once there, execute:

import Pkg; Pkg.add("Clapeyron")
import Pkg; Pkg.add("PyCall")

This installs the Clapeyron library and PyCall, which is necessary for calling Julia from Python.

Create a Python virtual environment to work comfortably and execute from the terminal:

    pip install julia

Then, in Python:

    import julia
    julia.install()

This installs PyJulia.

Now, calling Julia from Python won't work due to incompatibilities. The solution is to create a system image where Julia is compiled with Clapeyron and PyCall. See the documentation at: https://pyjulia.readthedocs.io/en/latest/sysimage.html

Execute from the library's root: python3 -m julia.sysimage psrk/sys.so

This command takes some time to run. It creates a compiled version of Julia and saves it in a sys.so file (system image).

I will create a sys.so in the psrk folder within this library so that it can be used directly. However, it may be necessary to create one on each machine because the compiled version may be different, even if the

To check if it works and learn a bit more, refer to the file calling_julia.ipynb.

Install the library in the virtual environment:
    pip install .

Optional:
you can run the tests to see if all works well.

pip install -r requirements_dev.txt

tox -r

Ejecutar desde bash:  
python3 -m julia.sysimage sys.so

SalvadorBrandolin/psrkwrap