Point cloud in CSV file as output (only data driven)

[matiyanez]

I have a set of 3d CFD simulations of a seawater intake, in which I have parameterized the speed of the seawater current and the suction speed of the seawater intake, from the simulations I have extracted a cloud of points with position information (x, y, z) and its respective velocity magnitude for each simulation, this in CSV format.
Each point cloud information is in a separate csv file.

I want to train a neural network with two input variables (the speeds that I have mentioned) and as output variable a cloud of points with the variables indicated above (x, y, z) and speed magnitude.

My question is how to tell modulus that my output is a csv file that contains the information of the cloud of points.

[ngeneva]

Hi @matiyanez

Yes you can make data-driven models inside of Modulus. CSV is a great format for the data to be in.

How to create a digital twin with only data-driven?

Hi @matiyanez Yes, Modulus does allow for physic-driven, hybrid, and data-driven training. An example of a pure data-driven problem is the Darcy problem which has image like data. For point wise data you can use the PointwiseConstraint.from_numpy() where you can feed in a set of numpy dictionaries to train from. This is used in a couple spots of our examples such as the three fin heat sink which can combined with other physics-based training constraints.

Reading from a CSV is done in a lot of our examples for reference, such as the cylinder flow problem.