Sample Nearest Points

PCGEx | Sample : Nearest Point

Sample points within a spherical range

The Sample Nearest Points node extracts and blends data from nearby target points within a customizable range, allowing you to fine-tune sampling methods, apply filters, and compute weighted outputs, making it ideal for tasks like proximity-based data collection.

Points that will sample data from targets

Point Filters

Points filters used to determine which points will be processed. Filtered out points will be treated as failed sampling.

Targets

Target points to read data from

Out

In with extra attributes and properties

The Sample Nearest Point grabs and blends attributes & properties from a target dataset, as well as other spatial relationship outputs.

details/sampling-nearest-point/lead.png

Sampling

Property	Description
Settings
Sample Method	Selects the sampling method. See Sampling Methods.
Range Min	Minimum sampling range.
Range Max	Maximum sampling range. Use `0` to sample all targets.
Local Range Min	If enabled, uses a per-point `double` attribute value as minimum sampling range.
Local Range Max	If enabled, uses a per-point `double` attribute value as maximum sampling range.

Points that are not within range are ignored. If no point is found within the specified range, the sampling for that point will be marked as Usuccessful.

Distance Details
Source	TBD
Target	TBD

Weighting
Weight Method	Selects the method used to compute the weight of each target. See Weighting.
Weight Over Distance	Curve used to sample the final weight of each target.

Sampling Methods

Method	Description
All (Within Range)	Samples all points within the specified range.
Closest Target	Sample the single closest target within the specified range.
Farthest Target	Sample the single farthest target within the specified range.

Weighting

details/sampling-nearest-point/weighting.png

There is two weighting method available. Full Range and Effective Range. Each method outputs a [0..1] value that will be used to sample the Weight Over Distance curve.
However, there is a critical nuance between the two:

Full Range is a simple normalization, each target distance is divided by the longest one. As such, it’s very unlikely the curve will get sampled close to x=0.
Effective Range remaps each target distance using the shortest & longest distance as min/max. As such, the shortest sampled distance will sample the curve at x=0, and the longest at x=1.

Note that the Effective Range method tends to spread/scale the input set of values – but allows one to leverage the full range of the curve no matter the min/max input values.
Hence, using Full Range with only high (or low) input value will only sample a very narrow portion of the curve.

Important note: when using the Within range sample method, some outputs will use the final weighted position/transforms for their calculations; although mathematically correct, this may yield unusuable/innacurate results.

Blending

Properties	Description
Target Attributes	A `Name`::`Blendmode` map of the attributes to grab from the target points, and which blendmode to use.
Point Propeprties Blending Settings	If enabled, lets you pick & choose which point properties you want to grab from the targets, and which blendmode to use. Use `None` to prevent a specific property from being grabbed.

See Blending.

Outputs

Outputs are values extracted from the neighbor(s), and written to attributes on the output points.

Output	Description
Generic
Success `bool`	TBD

Spatial Data
Transform `FTransform`	TBD
Look At `FVector`	TBD
└─ Align	TBD
└─ Use Up from…	TBD
└─ Up Vector	TBD
Distance `double`	TBD
Signed Distance `double`	TBD
└─ Axis	TBD
Angle `double`	TBD
└─ Axis	TBD
└─ Range	TBD
Num Samples `int32`	TBD

Based on the selected Sample method, the output values are a weighted average of all the sampled targets. See Weighting.

Tagging

Some high level tags may be applied to the data based on overal sampling.

Tag	Description
Has Successes Tag	If enabled, add the specified tag to the output data if at least a single target has been sampled.
Has No Successes Tag	If enabled, add the specified tag to the output data if no target has been sampled.

Note that fail/success tagging will be affected by points filter as well; since filtered out points are considered fails.

Advanced

Property	Description
Process Filtered Out As Fails	If enabled, mark filtered out points as “failed”. Otherwise, just skip the processing altogether. Only uncheck this if you want to ensure existing attribute values are preserved. Default is set to true, as it should be on a first-pass sampling.