What Is It?
A Surface Plot creates a continuous surface from your data in either 2D or 3D to view how one feature is affected by two others.
Why Is This Important?
Surface plots have specific use cases where they can be critical to understanding a problem. Most commonly, Surface plots are used to view the implied volatility surface, where you visualize how implied volatility of stock options is affected by time to maturity and strike price. You could also use a 2D Surface plot as a heatmap to view temperature as it changes over two dimensions of time (such as week and year).
Please note that to be able to create a Surface plot, your dataset should be structured so that you have one feature which represents a function of two other features, or Y = f(X, Z).
To ensure this requirement, check that the total number of points in your dataset equals the number of unique values in X multiplied by the number of unique values in Z.
The Group By dimension can also be used if there are multiple surfaces described in the data. In this case, each category in the feature mapped to Group By must individually satisfy the above criteria.
Steps for creating a Surface plot:
- Drag one feature from the Features list to the X dimension and one feature to the Z dimension (please note that these should be the X and Z features mentioned in the function above).
- If you click Apply, you will create a 2D Surface plot, the most basic option. This will show a single solid surface.
- By adding a feature to the Color dimension (please note that this should be the Y feature mentioned in the function above), you will create a heatmap that shows how your Color feature is affected by the X and Z features.
- Typically, a Color gradient looks best on this plot. You can change from Color bins to a Color gradient by opening the Color settings.
- Adding the same feature that is on Color to the Z dimension will create a 3D Surface plot so that you can view the relationships between your features in three dimensions.