Repository navigation
node normalize
< Previous: List Length | Next: Reverse >
Scale a list of values into 0 to 1 - the smallest becomes 0, the largest becomes 1. A list with no spread comes out all zeros.
Category: Data | Add menu: Data > Normalize
The smallest value in the list becomes 0, the largest becomes 1, and everything else lands where it sits between them. The range comes from the list itself, which is the whole point: it is how a list of measured lengths, areas or distances becomes something a Field, a scale or a color ramp can take without knowing what the numbers meant.
Order and length are preserved, so the result still pairs index for index with whatever the values came from.
- Values (number): the list to scale.
- Values (number): the same list, on a 0 to 1 footing.
This is why it sits under Lists rather than beside Math. The answer for one item depends on every other item: add one long value and all the rest restretch. Every other numeric node in the category answers index for index, and this one cannot.
It is the complement of Series. Series makes an index sequence; this takes whatever a graph has measured and puts it on the same footing.
- A list with no spread comes out all zeros. Every value being the same is the divide-by-nothing case: there is no position within a range of width zero to report. Zero rather than an error is the same guard Math's divide, modulo and power already use.
- The test is near zero, not exactly zero, so a span of a thousandth of a millionth does not get scaled up into a full 0 to 1 and report differences that are not there.
- A single value, not a list, is the same case for the same reason, and comes out 0.
- A value that is not a finite number reads as 0 and counts toward the range, rather than poisoning the whole answer.
- One level of list. Normalize reads a flat list of numbers; it does not reach inside a nested one.
Vector Math has an operation called Normalize too, and it is a different thing: it scales one vector to unit length, direction kept. This node scales a list of numbers into 0 to 1. Nothing is shared between them but the word.
- Series: the index sequence this is the complement of.
- List Length: how many items a list holds.
- Graph Mapper: reshaping values once they are on a 0 to 1 footing.
- Vector Math: the other Normalize, on a single vector.
- Data nodes: the rest of the category.
< Previous: List Length | Next: Reverse >