Normalization transforms the points along a dimension so that you can better see their distribution. It spreads out a tightly clustered group of points so that the differences are clearer in your visualization. It is often used to adjust values measured on different scales to a notionally common scale.
Normalization can be set in four ways:
- The Normalization button in the toolbar, which allows you to apply the same normalization to all three axes at once
- Normalization dropdown for each individual dimension in the Settings tab ( XYZ, Color, Others) on the Mapping panel
- The right-click contextual menu for a numerical feature in the Feature panel (the Normalize option lets you create new features that are normalized)
- This can also be done via the right-click Contextual Menu
Virtualitics supports the following different types of normalization:
- Log10 – The standard normalization type, log transformation changes the feature from linear to logarithmic. Cannot be used on features that contain a value of 0 or less.
- IHST – Similar to Softmax, but can provide slightly more space between points for certain features
- Softmax – Useful for normalizing features that contain values of 0 or less
- CDF – Cumulative distribution function