Skip to content

modelflows/ModelFLOWs-app

Repository files navigation

ModelFLOWs-app

Welcome to the ModelFLOWs-app repository!

Table of contents

About ModelFLOWs-app

ModelFLOWs-app, an open source Software for data post-processing, patterns identification and development of reduced order models using modal decomposition and deep learning architectures.

Versions

There are currently two versions available:

  • ModelFLOWs-app desktop version
  • ModelFLOWs-app web-browser version (demo)

Requirements

ModelFLOWs-app has been developed with:

  • Python: 3.9

Other relevant libraries that require installation:

  • tensorflow: 2.10
  • scikit-learn: 1.2
  • ffmpeg: latest version
  • hdf5storage: latest version
  • numba: 0.56.4
  • scipy: 1.9.3
  • keras-tuner: latest version
  • protobuf: 3.20

Setup

There are two ways to install all required libraries:

Option 1. Install all libraries:

$ cd ../v0.1_ModelFLOWs_app
$ sudo pip install -r Requirements.txt

Option 2. Install each library individually:

$ cd ../Desktop
$ sudo pip install **insert library name**

Run

To open ModelFLOWs-app, run the following command:

For the desktop version:

$ cd ../v0.1_ModelFLOWs_app
$ python ModelFLOWs_app.py

For the web-browser demo:

IMPORTANT Before running this version, please read the instructions that appear in the web-browser-version branch to download the databases used by the ModelFLOWs-app web-browser version. If the databases aren't downloaded correctly, the web-browser version will not work.

$ cd ../v0.1_ModelFLOWs_web
$ streamlit run ModelFLOWs_web.py

Resources

Check out the Tutorial folder for a more in depth tutorial explaining how to install, run and use ModelFLOWs-app. This folder also contains advanced tutorials explaining how each module of the application works.

More info

For more information please visit ModelFLOWs-app's official website. You can also find us on LinkedIn