diff --git a/SFEM/README.md b/SFEM/README.md index 881b21e370e84799c18f8ca8b3aa38d51b04071a..84ebb92dd849ba15b4da3addc22441551c653130 100644 --- a/SFEM/README.md +++ b/SFEM/README.md @@ -19,7 +19,7 @@ This repository contains: ## Dependencies and Prerequisites -[](https://www.python.org/downloads/) [](https://github.com/pandas-dev/pandas) [](https://matplotlib.org/) +[](https://www.python.org/downloads/) [](https://github.com/pandas-dev/pandas) [](https://pytorch.org/get-started/locally/#linux-pip) [](https://matplotlib.org/) * Python, pandas, matplotlib, texttabble and latextable are pre requisites for visualizing and navigating the data. @@ -42,6 +42,11 @@ Instructions for Debian/Ubuntu based workstations are as follows. ``` bash pip3 install matplotlib texttable latextable ``` +### Pytorch (only for run with cm3Libraries) + +``` bash + pip3 install --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu +``` ## Structure of Repository @@ -71,12 +76,27 @@ Instructions for Debian/Ubuntu based workstations are as follows. * [plotRF.py](./generateRF/plotRF.py): file used to vizualize the distribution of the random variables. * Reads the ```N``` random fields of type ```'RandField_X.csv'``` found in ```mechDir+'/randomFields/'```, where ```mechDir``` is the working directory. Examples of generated 2-dimension RFs can be found in [rnnRF/randomFields/](./rnnRF/randomFields/). * Plot the distribution of the random variables. -* [vizualizeRF.py](./generateRF/vizualizeRF.py): file used to generate a gmsh compatible file to vizualize the random fields +* [vizualizeRF.py](./generateRF/vizualizeRF.py): file used to generate a gmsh (www.gmsh.info) compatible file to vizualize the random fields * Reads the ```N``` random fields of type ```'RandField_And_GP_X.csv'``` found in ```mechDir+'/randomFields/'```, where ```mechDir``` is the working directory. Examples of generated 2-dimension RFs can be found in [rnnRF/randomFields/](./rnnRF/randomFields/). * Uses the mesh structure of the considered application. The script tries to read stress file in ```mechDir+'/GPData'```, where ```mechDir``` is the working directory, describing the elements structure. One example of stress file can be found in [rnnRF/GPData/stress_step1.msh](./rnnRF/GPData/stress_step1.msh). * The random field number ```X``` is saved in format compatible with Gmhs in the file ```mechDir+'/randomFields/RandField_X.msh'```, where ```mechDir``` is the working directory. One example of generated 2-dimension RF can be found in [rnnRF/randomFields/RandField_0.msh](./rnnRF/randomFields/RandField_0.msh). * [utilRF.py](./generateRF/plotRF.py): set of functions used by the RF generator and vizualization files. +### rnnRF +* [rubics.geo](./rnnRF/rubics.geo): geometry file of a cube to be read by gmsh (www.gmsh.info). +* [rubics.msh](./rnnRF/rubics.msh): mesh file of a cube generated by gmsh (www.gmsh.info). +* [macro_cpp.py](./rnnRF/macro_cpp.py): multiscale lattice compression script to be run with cm3Libraries (http://www.ltas-cm3.ulg.ac.be/openSource.htm) and using a [MOAMMM stochastic neural network](https://gitlab.uliege.be/moammm/moammmPublic/syntheticdata/sveresponses) as surrogate of the cell response. +* [Bounds.dat](./rnnRF/Bounds.dat): Bounds of the neural network model. +* [model.dat](./rnnRF/model.pt): torch neural network model. +* [generateGPInfo.py](./rnnRF/generateGPInfo.py): generate the Gauss points coordinates and mesh structure to be used by the random field generator. It requires cm3Libraries (http://www.ltas-cm3.ulg.ac.be/openSource.htm). + * The script write the Gauss points coordinates in the directory ```'rnnRF/GPData'```. One example of points coordinates can be found in [rnnRF/GPData/coordinate_P_ZZ_OnPhysical_11.csv](./rnnRF/GPData/coordinate_P_ZZ_OnPhysical_11.csv) stored following 3 coordinates, point number and point volume (for a FE simulation). + * The script write the mesh structure in the directory ```'rnnRF/GPData'```. One example of stress file can be found in [rnnRF/GPData/stress_step1.msh](./rnnRF/GPData/stress_step1.msh). +* [runMCTest.py](./rnnRF/runMCTest.py): Monte Carlo simulations on a multiscale lattice compression using a [MOAMMM stochastic neural network](https://gitlab.uliege.be/moammm/moammmPublic/syntheticdata/sveresponses) as surrogate of the cell response. It requires cm3Libraries (http://www.ltas-cm3.ulg.ac.be/openSource.htm). + * The script uses the ```N``` random fields of type ```'RandField_And_GP_X.csv'``` found in ```rnnRF/randomFields/'```. Examples of 2-dimension RFs can be found in [rnnRF/randomFields/](./rnnRF/randomFields/). + * The script saves the simulations results in ```'rnnRF/results/RandField_And_GP_X.csv/'``` +* [plotMC.py](./rnnRF/plotMC.py): vizualize the different force displacement curves of the MC simulations. + * The script reads the simulations results saved in ```'rnnRF/results/RandField_And_GP_X.csv/'``` + ### cellRF