AIROYoung_Tutorial

In this repository there are easy examples or exercise concerning the main functionalities of Boost Python.

Boost Python is a library in the Boost suite which allows a very simple way to interface your C++ code to Python.

Structure

The tutorial is structured as follows:

    AYT_wd
    ├───Exercise
    │   ├───build
    │   │   └───source
    │   └───python_scripts
    │   ├───data
    ├───Exercise_Solved
    │   ├───build
    │   │   └───source
    │   └───python_scripts
    │   ├───data
    ├───Tutorial
    │   ├───build
    │   │   └───source
    │   └───python_scripts
    └───utils

1. In the Tutorial folder there is the (Python and C++) source code for a Python package to model an example concerning all the main Boost Python instruction for base and medium user;

2. In the Exercise folder there is the the (Python and C++) source code for a Python package to model an exercise with empty implementation (TODO) of methods for the exposition to Python;

3. In the Exercise_Solved folder there is the same C++ source code of Exercise with the implementation.

4. In the utils folder there are some miscellaneous Python functions used by the notebooks

Given any of the previous folder, the inner structure is the same for the three of them:

    folder
    ├───build
    │   │ CMakeLists.txt
    │   │ build_all.bat
    │   │ build_all.sh
    │   └─── source
    └───python_scripts
    └───data

In the build folder there is a source subdirectory containing the C++ files to compile in a Python module through Boost Python. The file build_all.bat/sh (depending on the platform) calls the build system configuration file CMakeLists.txt and creates the Python package, then it copies it in the parent directory. These files must be customized by the client, in particular one must decide the configuration/build type (Debug or Release) and must set up the Boost root directory. The folder python_scripts and data contain respectively additional files written in Python and data loaded for example purpose in the notebooks in the root folder. Once the package is built and in the folder directory, one should start the JuPyter server. It is very important that the JuPyter server is launched in the AYT_wd folder.

Setup

The tutorial needs to install some Python required packages, so you may prefer to use a Python virtual environment by virtualenv. That said, you just need to install what follows:

• Boost C++ Libraries (>= 1.66);
• CMake (>= 3.14);
• Python (== 2.7.15).

Depending on some characteristics of the Boost libraries you may need to change some directives in the CMakeLists.txt, but they are very easy to understand:

#tags of boost libraries
SET(Boost_USE_STATIC_LIBS ON) #set to ON if you compiled boost as static libraries
SET(Boost_NO_SYSTEM_PATHS ON) #do not modify this
SET(Boost_USE_MULTITHREADED ON) #set to ON if you compiled boost with the multithreaded option
SET(Boost_ARCHITECTURE -x64) #set to ON if you compiled boost for x64 platform

If you have any troubles finding the Boost libraries, add to the CMakeLists.txt file the command set(Boost_DEBUG ON) for a more verbose output.

CMake will automatically find a C++ compiler, but in case of any troubles refer to the CMake documentation.

At this point, you just need the following steps to run the code:

1. [optional] make use of a python virtual environment;
2. execute pip install -r requirements.txt;
3. choose one package to build and execute the build_all.bat/sh file;
4. launch a JuPyter server from the AYT_wd folder ([optional] in the same virtual environment as in 1);
5. open the notebook and execute each cell one by one to analyze the output.