How to evaluate surface field predictions

Modified on Tue, 30 May 2023 at 08:03 AM

When you train surface field models you can visually assess its accuraty by using the step View Surface Field Prediction and looking at the true, predicted, and error fields. However, as an engineer you often have detailed questions concerning these fields: 

  • What is the maximum or minimum value occurring in the field?
  • What is my average error in the complete field?
  • What is the integral value over the entire field?

The surface field model offers no direct access to answer these questions, but the platform offers tools to derive the answer within a few clicks.

  • Use 3D Dataset Prediction to create a new dataset which contains the predicted fields (View Surface Field Prediction is not storing the predicted field in a new dataset).
  • Use 3D to Table to transform the surface field prediction into tabular data. On that you can use all tools for tabular data in the platform.
  • With Quick Columns you can calculate other variables from the surface field result.
  • With Group By you can extract scalar values from your field data.
    • The Max operation, for example, allows to extract the maximum value for each surface field prediction.
    • With Min, Mean and Sum you have several other possibilities to evaluate your surface field prediction.
    • As Group By and Quick Columns can both be used by writing your own Python code, the possibilities are countless (see here to know more about Python custom code).

1. 3D to Table and Group By enable to manipulate surface field predictions as tabular data.

2. Group By offers several operations on the data of the tabulated 3D field.

Was this article helpful?

That’s Great!

Thank you for your feedback

Sorry! We couldn't be helpful

Thank you for your feedback

Let us know how can we improve this article!

Select atleast one of the reasons
CAPTCHA verification is required.

Feedback sent

We appreciate your effort and will try to fix the article