Skip to content

Commit

Permalink
Merge pull request #267 from bessagroup/pr/1.5.1
Browse files Browse the repository at this point in the history 
Pr/1.5.1
  • Loading branch information
@mpvanderschelling
mpvanderschelling authored May 31, 2024
2 parents 38009a9 + d290d35 commit 67a21c6
Show file tree
Hide file tree
Showing 6 changed files with 10 additions and 157 deletions.
4 changes: 2 additions & 2 deletions README.md
View file
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ f3dasm
[![Documentation Status](https://readthedocs.org/projects/f3dasm/badge/?version=latest)](https://f3dasm.readthedocs.io/en/latest/?badge=latest)

[**Docs**](https://f3dasm.readthedocs.io/)
| [**Installation**](https://f3dasm.readthedocs.io/en/latest/rst_doc_files/general/gettingstarted.html)
| [**Installation**](https://f3dasm.readthedocs.io/en/latest/rst_doc_files/general/installation.html)
| [**GitHub**](https://github.com/bessagroup/f3dasm)
| [**PyPI**](https://pypi.org/project/f3dasm/)

Expand Down Expand Up @@ -40,7 +40,7 @@ Welcome to `f3dasm`, a **f**ramework for **d**ata-**d**riven **d**esign and **a*

* Read the [overview](https://f3dasm.readthedocs.io/en/latest/rst_doc_files/general/installation.html) section, containing a brief introduction to the framework and a statement of need.
* Follow the [installation instructions](https://f3dasm.readthedocs.io/en/latest/rst_doc_files/general/installation.html) to get going!
* Check out the [Tutorials](https://f3dasm.readthedocs.io/en/latest/auto_examples/index.html) section, containing a collection of examples to get you familiar with the framework.
* Check out the [tutorials](https://f3dasm.readthedocs.io/en/latest/auto_examples/index.html) section, containing a collection of examples to get you familiar with the framework.

## Illustrative benchmarks

Expand Down
2 changes: 1 addition & 1 deletion VERSION
View file
Original file line number Diff line number Diff line change
@@ -1 +1 @@
1.5.0
1.5.1
4 changes: 2 additions & 2 deletions docs/source/conf.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -27,8 +27,8 @@
project = 'f3dasm'
author = 'Martin van der Schelling'
copyright = '2024, Martin van der Schelling'
version = '1.5.0'
release = '1.5.0'
version = '1.5.1'
release = '1.5.1'


# -- General configuration ----------------------------------------------------
Expand Down
147 changes: 0 additions & 147 deletions docs/source/rst_doc_files/general/overview.rst
View file
Original file line number Diff line number Diff line change
Expand Up @@ -69,150 +69,3 @@ Comprehensive `online documentation <https://f3dasm.readthedocs.io/en/latest/>`_
.. [1] Bessa, M. A., Bostanabad, R., Liu, Z., Hu, A., Apley, D. W., Brinson, C., Chen, W., & Liu, W. K. (2017).
*A framework for data-driven analysis of materials under uncertainty: Countering the curse of dimensionality.
Computer Methods in Applied Mechanics and Engineering*, 320, 633-667.
.. The use of state-of-the-art machine learning tools for innovative structural and materials design has demonstrated their potential in various studies.
.. Although the specific applications may differ, the data-driven modelling and optimization process remains the same.
.. Therefore, the framework for data-driven design and analysis of structures and materials (:mod:`f3dasm`) is an attempt to develop a systematic approach of inverting the material design process.
.. The framework, originally proposed by Bessa et al. :cite:p:`Bessa2017` integrates the following fields:
.. - **Design \& Sampling**, in which input variables describing the microstructure, structure, properties and external conditions of the system to be evaluated are determined and sampled.
.. - **Simulation**, typically through computational analysis, resulting in the creation of a material response database.
.. - **Machine learning**, in which a surrogate model is trained to fit experimental findings.
.. - **Optimization**, where we try to iteratively improve the model to obtain a superior design.
.. The effectiveness of the first published version of :mod:`f3dasm` framework has been demonstrated in various computational mechanics and materials studies,
.. such as the design of a super-compressible meta-material :cite:p:`Bessa2019` and a spiderweb nano-mechanical resonator inspired
.. by nature and guided by machine learning :cite:p:`Shin2022`.
.. .. [3] Bessa, M. A., Bostanabad, R., Liu, Z., Hu, A., Apley, D. W., Brinson, C., Chen, W., & Liu, W. K. (2017).
.. *A framework for data-driven analysis of materials under uncertainty: Countering the curse of dimensionality.
.. Computer Methods in Applied Mechanics and Engineering*, 320, 633-667.
.. Modularity and use cases
.. ^^^^^^^^^^^^^^^^^^^^^^^^
.. The package contains a lot of implementation for each of the blocks.
.. However, the installation :mod:`f3dasm` is modular: you decide what you
.. want to use or not.
.. We can distinguish 3 ways of using :mod:`f3dasm`:
.. Using :mod:`f3dasm` to handle your design of experiments
.. -----------------------------------------------------
.. The :mod:`f3dasm` package: contains the minimal installation to use
.. :mod:`f3dasm` without extended features.
.. .. note::

.. You can install the core package with ``pip install f3dasm`` or `read the installation instructions <https://bessagroup.github.io/f3dasm/general/gettingstarted.html>`__!
.. The core package contains the following features:
.. 1. provide a way to parametrize your experiment with the `design-of-experiments`_ classes.
.. 2. provide the option to investigate their experiment by `sampling`_ and `optimizing`_ their design.
.. 3. provide the user guidance in `parallelizing`_ their program and ordering their data.
.. 4. give the user ways of deploying their experiment at a high-performance computer system (`TORQUE system`_).
.. The core package requires the following dependencies:
.. - `numpy <https://numpy.org/doc/stable/index.html>`_ and `scipy <https://docs.scipy.org/doc/scipy/reference/>`_: for numerical operations
.. - `pandas <https://pandas.pydata.org/docs/>`_ and `SALib <https://salib.readthedocs.io/en/latest/>`_: for the representation of the design of experiments
.. - `matplotlib <https://matplotlib.org/stable/contents.html>`_: for plotting
.. - `hydra-core <https://hydra.cc/docs/intro/>`_: for deploying your experiment
.. - `pathos <https://pathos.readthedocs.io/en/latest/>`_: for multiprocessing
.. - `autograd <https://github.com/HIPS/autograd>`_: for computing gradients
.. Using :mod:`f3dasm` extended capabilities
.. --------------------------------------
.. Use existing implementations to benchmark parts of the data-driven machine learning process!
.. For this purpose, you can solely use the core package, but it is advised
.. to enrich :mod:`f3dasm` with its **extension libraries**
.. The extensions contain the following features:
.. 1. provide various **implementations** to accommodate common machine learning workflows.
.. 2. provide **adapter** classes that link common machine learning libraries to :mod:`f3dasm` base classes.
.. The following extensions libraries are available:
.. - `f3dasm_simulate <https://github.com/bessagroup/f3dasm_simulate>`_: containing various simulators ported to be used with :mod:`f3dasm`.
.. - `f3dasm_optimize <https://github.com/bessagroup/f3dasm_optimize>`_: containing various optimizers from `GPyOpt <https://gpyopt.readthedocs.io/en/latest/>`_, `pygmo <https://esa.github.io/pygmo2/index.html>`_ and `tensorflow <https://www.tensorflow.org/api_docs/>`_
.. The main takeaway is that if your design-of-experiments is modified to
.. use the ``f3dasm.ExperimentData`` class, you are able to seamlessly
.. incorporate the extension into your application!
.. Abstraction
.. ^^^^^^^^^^^
.. By abstracting away the details of specific implementations, users and developers can better organize and reuse their code,
.. making it easier to understand, modify, and share with others. Within the :mod:`f3dasm` framework, abstraction is done in four levels:
.. - **block**: blocks represent one of the high-level stages that can be used in the framework, e.g. the :mod:`~f3dasm.optimization` submodule. They can be put in any specific order, and incorporate a core action undertaken by the design.
.. - **base**: bases represent an abstract class of an element in the block, e.g. the :class:`~f3dasm.optimization.optimizer.Optimizer` class. Base classes are used to create a unified interface for specific implementations and are inherited from blocks.
.. - **implementation**: implementations are application of a base class feature, e.g. the :class:`~f3dasm.optimization.adam.Adam` optimizer. These can be self-coded or ported from other Python libraries.
.. - **experiment**: experiments represent executable programs that uses a certain order of blocks and specific implementations to generate results.
.. .. image:: ../../img/f3dasm-blocks.svg
.. Overview of implementations and base classes
.. --------------------------------------------
.. ===================== =============================== ========================================================================== =======================================================
.. Block Submodule Base Implementations
.. ===================== =============================== ========================================================================== =======================================================
.. Design of Experiments :mod:`~f3dasm.design` :class:`~f3dasm.design.design.Domain`
.. :mod:`~f3dasm.sampling` :class:`~f3dasm.sampling.sampler.Sampler` :ref:`List of samplers <implemented samplers>`
.. Data generation :mod:`~f3dasm.datageneration` :class:`~f3dasm.datageneration.DataGenerator` :ref:`List of datagenerators <implemented datagenerators>`
.. Machine learning :mod:`~f3dasm.machinelearning` :class:`~f3dasm.machinelearning.model.Model` :ref:`List of models <implemented models>`
.. Optimization :mod:`~f3dasm.optimization` :class:`~f3dasm.optimization.optimizer.Optimizer` :ref:`List of optimizers <implemented optimizers>`
.. ===================== =============================== ========================================================================== =======================================================
.. Overview of other classes
.. -------------------------
.. =============================================================== ===================================================================================
.. Class Short description
.. =============================================================== ===================================================================================
.. :class:`~f3dasm.ExperimentData` Datastructure denoting samples from a design-of-experiments
.. :class:`~f3dasm.functions.function.Function` Class that represents an analytical function used for benchmarking
.. :class:`~f3dasm.functions.adapters.augmentor.Augmentor` Class that can be used to manipulate data for data-augmentation
.. :class:`~f3dasm.optimization.optimizer.OptimizerParameters` Class that represents the hyper-parameters for a particular optimizer
.. :class:`~f3dasm.run_optimization.OptimizationResult` Class used to store optimization results for several epochs
.. =============================================================== ===================================================================================
.. References
.. ----------
8 changes: 4 additions & 4 deletions paper/paper.md
View file
Original file line number Diff line number Diff line change
@@ -1,5 +1,5 @@
---
title: 'f3dasm: Framework for Data-Driven Design & Analysis of Structures & Materials'
title: 'f3dasm: Framework for Data-Driven Design and Analysis of Structures and Materials'
tags:
- Python
- data-driven
Expand All @@ -10,7 +10,7 @@ authors:
- name: M. P. van der Schelling
orcid: 0000-0003-3602-0452
affiliation: 1
- name: B.P. Ferreira
- name: B. P. Ferreira
orcid: 0000-0001-5956-3877
affiliation: 2
- name: M. A. Bessa
Expand All @@ -29,7 +29,7 @@ bibliography: paper.bib

# Summary

[`f3dasm`](https://github.com/bessagroup/f3dasm) (Framework for Data-driven Design and Analysis of Structures \& Materials) is a Python project that provides a general and user-friendly data-driven framework for researchers and practitioners working on the design and analysis of materials and structures. The package aims to streamline the data-driven process and make it easier to replicate research articles in this field, as well as share new work with the community.
[`f3dasm`](https://github.com/bessagroup/f3dasm) (Framework for Data-driven Design and Analysis of Structures and Materials) is a Python project that provides a general and user-friendly data-driven framework for researchers and practitioners working on the design and analysis of materials and structures. The package aims to streamline the data-driven process and make it easier to replicate research articles in this field, as well as share new work with the community.

![Logo of [`f3dasm`](https://github.com/bessagroup/f3dasm). \label{fig:f3dasm_logo}](f3dasm_logo_long.png)

Expand All @@ -39,7 +39,7 @@ In the last decades, advancements in computational resources have accelerated no

Constructing a large material response database poses practical challenges, such as proper data management, efficient parallel computing, and integration with third-party software. Because most applied fields remain conservative when it comes to openly sharing databases and software, a lot of research time is instead being allocated to implement common procedures that would be otherwise readily available. This lack of shared practices also leads to compatibility issues for benchmarking and replication of results by violating the FAIR principles.

In this work we introduce an interface for researchers and practitioners working on the design and analysis of materials and structures. The package is called [`f3dasm`](https://github.com/bessagroup/f3dasm) (Framework for Data-driven Design \& Analysis of Structures and Materials). This work generalizes the original closed-source framework proposed by the Bessa and co-workers [@Bessa2017], making it more flexible and adaptable to different applications, namely by allowing the integration of different choices of software packages needed in the different steps of the data-driven process:
In this work we introduce an interface for researchers and practitioners working on the design and analysis of materials and structures. The package is called [`f3dasm`](https://github.com/bessagroup/f3dasm) (Framework for Data-driven Design and Analysis of Structures and Materials). This work generalizes the original closed-source framework proposed by the Bessa and co-workers [@Bessa2017], making it more flexible and adaptable to different applications, namely by allowing the integration of different choices of software packages needed in the different steps of the data-driven process:

- **Design of experiments**, in which input variables describing the microstructure, properties and external conditions of the system are determined and sampled;
- **Data generation**, typically through computational analyses, resulting in the creation of a material response database [@Ferreira2023];
Expand Down
2 changes: 1 addition & 1 deletion src/f3dasm/__version__.py
View file
Original file line number Diff line number Diff line change
@@ -1 +1 @@
__version__: str = "1.5.0"
__version__: str = "1.5.1"

0 comments on commit 67a21c6

Please sign in to comment.