snoopy-bv 1.3.2

DR C++/Python library

Snoopy library

Documentation and tutorials : https://bv-dr.gitlab.io/Snoopy/

Snoopy is an open source library for marine applications. Among other things, it contains the following sub-packages

• Spectral : Routines for spectral analysis of sea-keeping results

  • Wave spectra (Jonswap, Ochihubble...)
  • Response spectrum
  • 2nd order response spectrum

• TimeDomain

  • Wave kinematic model (Linear, wheeler stretching, 2nd order)
  • Reconstruction of seakeeping results in time domain (1st and 2nd order)
  • Convolution for wave radiation in time domain (from added-mass).

• Fatigue

  • Rainflow counting
  • Spectral fatigue

• Meshing

How to install from pre-compiled library

Pre-compiled wheel package can be installed with :

pip install snoopy-bv --no-deps

or, to upgrade (as version number is not always changed) :

pip install snoopy-bv --force-reinstall

In case conda is used, it is recommended to install dependencies first, using conda:

conda install --file requirements.txt

How to build from sources

• Create a build folder, for instance "build"

• Run cmake

go in "build" directory and run :

cmake .. -G"CodeBlocks - MinGW Makefiles"

• Compile

mingw32-make.exe -j10

How to use in develop mode

Python should know how to find the compiled pyd (or .so on linux). Two solutions :

• Add the folder containing the .pyd to the PYTHONPATH
• Add the folder containng the .pyd in the SNOOPY_PYD environement variable

To install in "developer mode" (use current sources, do not copy in python environment) :

pip install -e C:\path_to_snoopy_location

How to install

to install in python environment :

pip install pathToRootfolder

How to package

• The folder containing the .pyd should be in the SNOOPY_PYD environement variable

python setup.py bdist_wheel

How to install on Linux (tested with Ubuntu 20.04 with bash terminal)

• clone in the folder snoopy_dev (using snoopy as name may cause problems)

git clone https://mar-gitlab.bureauveritas.com/applications/dr/snoopy.git snoopy_dev

• compile the code in a build directory

cd snoopy_dev/Snoopy
mkdir build
cd build
cmake ..
make -j 10

• install the package (first export the variable SNOOPY_PYD to the path to the compiled libraries)

cd ../..
export SNOOPY_PYD=./build/lib
pip install .

To install in develop mode:

pip install -e .

Dev notes:

Commit message

• [FUN] : New feature
• [ENH] : Enhancement of existing code
• [BUG] : Bug correction, [BBUG] for really BIG BUG (with consequences in case of use...)
• [QLT] : Quality (Comments, removal of dead code...)
• [FIX] : Correction of compiling issue
• [DAT] : Data added or modified
• [DOC] : Document added or modified
• [TEST] : Test added or modified
• [INI] : Initial commit.

The subpackage should be mentioned after this Flag. For instance :

[ENH] TimeDomain : Performance optimization of ReconstructionRaoLocal