Welcome to the Inductiva API Tutorials

Welcome to the Inductiva API Tutorials#

Here you will find in-depth guides explaining the inner workings of the Inductiva API,as well as how to use it for achieving certain pratical goals. These tutorials add to the information snippets available in the official API documentation by providing more detailed step-by-step explanations and instructions.

Available Tutorials#

  • Introduction to Inductiva API. In this tutorial, we will give you a comprehensive overview of how the API works, explaining key concepts and how different components play together. If you never used the API before, we recommend you read this tutorial.

Inductiva API Usage Flow

  • Generating Synthetic Data for training Physics-ML models. Synthetic data allows us to train Physics-ML models when you don’t have enough observational data, which often happens in many practical scenarios. In those cases, we can use physical simulators to mimic real-world dynamics under various simulated conditions, and produce data that can help the model to learn the specifics of the scenario while respecting the underlying physical laws. This tutorial series will walk you through the process of using the Inductiva API for generating one of such synthetic training dataset to enable your ML models to learn complex fluid dynamics.

  • Partial Differential Equations – Finite Differences and Physics-Informed Neural Networks. A step by step tutorial on numerical methods to solve Partial Differential Equations, where we use the 2D Heat Equation as our working example. We start with the Finite-Differences method, introduce the main concepts behind Physics-Informed Neural Networks, and conclude with a Generalized Neuro-Solver that can handle more complex geometries and varying initial/boundary conditions;