Collection of Codes

Source Code is freely available to use the GNU Public License.

All codes!

ENS-Challenge (dreem)

Here is available the ipynb for the dreem ENS-Challenge at College-de-France, for the course of Modèles Multiéchelles et Réseaux de Neurones Convolutifs (Prof. Sthephane Mallat) Code – Update: 23/03/2020 Some important remarks on the implementation: The ipynb was tested on a Docker container with TensorFlow (v2.1.0).

Python Implementations

This collection contains all implementation within the Engineering Degree Thesis. They are organized regarding the type (Elastic and Viscous cells problems, Elastodynamic models, Mesh Generation).

Cell Problems

They are described by ipynb files each case.

• 2-Dimensional Models

  • Modelling the cell problem using square type microstructure, with circular inclusion defining the porosity. Code

  • Modelling the cell problem using hexagonal type microstructure, with circular inclusion defining the porosity. This case defines the same microstructure as proposed by Parnell and Grimal reference studies. Code

• 3-Dimensional Model

  • Modelling the cell problem using a cubic type microstructure, with cylindrical inclusion defining the porosity level. This case seek the interaction between axial and non-axial mechanical behavior, just studying possible effects from non-axial component that modify the elastic behavior of bone. Code

Viscous Problems and quality factors

File is described by ipynb notebook. In this case, the homogenized elastic and viscous coefficients are deduced from a unitary square microstructure following the two-scale asymptotic heuristic. Such description is derived towards a separation between real and imaginary parts (associated to elastic and viscous parts) to assess and define the so-called quality factors. Code.

Elastodynamic models

Files are separated regarding fully elastodynamic models without attenuation and attenuated models. Different models are proposed, in time and frequency domains using the asymtotic homogenization scheme. SOON UPDATES!

Mesh Generation

Following the same trend as before, a schematic mesh generation is implemented on ipynb files, using octave kernel in combination with iso2mesh library. In this case, on a ipynb it is assessed the mesh generation from bmp cortical bone images after resizing (to be able to use on personal computers). It describes a full procedure to process images from computational tomography stacks, generate the volumetric image associated to the stack and mesh it using iso2mesh library. Code

Moreover, file convertion to XML file for FEniCS usage is also implemented for tethaedral meshes The python implementation, creates useful XML file format enconding tethraedral meshes with possibly subdomains if it is the case. It requires as input the XYZ coordinate matrix of vertices and connectivity matrix related to the XYZ points. Code XYZ to Mesh. Similarly, generation of meshes with tagged subdomains is implemented. Code XYZ to Subdomains

People think focus means saying yes to the thing you’ve got to focus on. But that’s not what it means at all. It means saying no to the hundred other good ideas that there are. You have to pick carefully. I’m actually as proud of the things we haven’t done as the things I have done. Innovation is saying no to 1,000 things.

Steve Jobs — Apple Worldwide Developers’ Conference, 1997