Class  Description 

AddTrainingDataFE  
AffineMapFE 
Implements an affine map  that is, y = Mx + b
where x is the input (a column vector), b is an additive constant vector,
M is a b*x matrix, and y is the output vector.

AngleBetweenVectorsFE 
Compute the interior angle between two vectors in degrees.

AngleBetweenVectorsFE.Signed2D  
AspectAngleFE 
Derive the target aspect angle, e.g.

AspectAngleFE.Interior01 
This override returns the unsigned interior angle in the range 0 to 1.
e.g. 
AspectAngleFE.InteriorDegrees 
This override returns the unsigned interior angle in degrees in the range 0 to 180
e.g. 
BetweenFE  
BetweenFE.Copy  
BetweenFE.DummyVar  
BetweenFE.Filter  
BetweenFE.Segment  
BufferObservationsFE 
BufferObservationSequencesFE collects the MLObservation Sequences from all problem instances
into a List

CentroidFE  
CentroidFE.LatLon  
ChaserFE 
Chaser is a very simple 2d physical simulation of an particle that
'chases' its input feature.

ClampFE  
ClosestPointOfApproachFE 
Project the closest point of approach for two points A and B assuming constant velocity for each.

ClosestPointOfApproachFE.Result  
ClusterPathsFE 
Cluster Paths is a batch processor.

CogFoundryFE 
CogFoundryFE implements an RBB ML Feature Extractor using a Cognitive Foundry Evaluator.

ConstantFE 
Output a constant value.

CreateEventFE 
Create a new RBB Event (discarding any timeseries data) for each problem instance.

CreateTimeseriesFE 
See helpString below

DecimateFE 
DecimateFE propagates the first and last observations, but
discards all others except each multiple of n.

DefaultValueFE 
Outputs the input unless the input is null, in which case a specified value is output.

DefineTimeCoordinate  
DefineTimeCoordinate.EndScale  
DefineTimeCoordinate.StartEnd  
DefineTimeCoordinate.StartScale  
DelayObservationsFE 
This is a base class for feature extractors that need to see all the results of their inputs
before computing their outputs, perhaps even before determining what their outputs will be.

DerivativeFE  
DimensionalityFE  
DistanceFE 
Computes the distance between pairs of inputs
or successive values of the same input (if the 1input constructor is used)
Outputs null if either of these values is null
Output:
y = a  b, the vector difference, if type = VECTOR
y = ab, the euclidean distance, if type = SCALAR
y = scalar distance in meters between locations given by latitude/longitude in decimal degrees, if type = SCALAR_LAT_LON_M
y = interior angle in degrees, if
If one of the inputs has lower dimension than the other (e.g.

DTWPathFE 
See the helpString below.

DTWSegmentationFE 
See helpString

DTWSegmentationFE.Demo  
DurationFE  
ExcelFE 
ExcelFe writes the batch as an excel workbook with a sheet for each problem instance.

ExecFE 
This feature extractor executes a process once for each problem instance.

FeedbackFE 
FeedbackFE creates a feature from the previous value of another feature.

FillNullsFE  
FollowerFE 
Project a location behind another point (the leader), given the leader's position and velocity.

FromToFE 
FromToFE creates a timeseries for each time the
individual input 'mover' moves from within a threshold distance of
some element of the group 'from' to within a threshold distance of
a different element of the group 'to'.

FutureValueFE 
Insert a feature that is a past or future value of another feature.

GaussianFE  
GaussianFE.MaxOfOne  
GnuplotFE 
Plots the data from problem instances.

GnuplotFE.Plotter  
GroupCountFE 
Outputs the number of members in the group

GroupCountFE.Current  
GroupCountFE.Total  
GroupLeaderFE  
GroupMeanFE  
GroupOutlierRemovalFE  
GroupPercentileFE  
IntegralFE 
Compute the integral of a singledimensional feature.

JavaScriptFE 
This feature extractor supports executing javascript code to process other
inputs.

MagnitudeFE 
Computes the magnitude of a feature, i.e.

MaximumFE 
Passes observation down the chain only once for each problem instance 
depending on the specific subclass used.

MedianFilterFE  
MinimumFE 
Passes observation down the chain only once for each problem instance 
depending on the specific subclass used.

NearestFE 
Creates a feature which is the position of the member of the group that
is nearest the reference point.

NearestFE.WithTags  
NearFE 
See NearFE.helpString for documentation.

NearFE.LatLon  
ObservationPlotFE  
ObservationRateFE  
ObserveOnceFE 
Passes observation down the chain only once for each problem instance,
however the init/done times are unchanged, so the single Observation stands
in for the entire sequence.

ObserveOnceFE.AtTime 
Use only the first observation at or after the specified time.

ObserveOnceFE.First  
ObserveOnceFE.Last 
Use only the final observation

ObserveTimespanFE 
Passes observation down the chain only between the specified start and end times.

OutnumberedFE 
See NearFE.helpString for documentation.

PassFailFE 
This is a base class for Feature Extractor classes that will admit an entire
timeseries if any of the observations pass some test.

PassFailFE.BBox 
Base class for gates that pass or fail based on an ndimensional bounding box.

PassFailFE.EndsIn 
Accept the observation sequence only if the first lastnull sample is inside the specified bounding box.

PassFailFE.EntersArea 
Accept timeseries that (at some point) go into a specified (hyper) rectangle
Any of the min or max extents can be null, in that case it is unbounded in that direction.

PassFailFE.ExcludeTags 
Exclude the observation sequence if any of the inputs have a specified tagset

PassFailFE.ObservationGap 
Accept observation sequence if the time between any pair of successive observations exceed the specified time threshold.

PassFailFE.StartsIn 
Accept the observation sequence only if the first nonnull sample is inside the specified bounding box.

PercentileFE 
Observes the specified percentile value of the specified feature in the final observation.

PreviousValueFE 
The current value of the output is a previous value of the specified input,
as
Computes the net change (dy) in one feature (y) as another
feature (x) increased by a specified amount (dx).

PrintCSVFE  
PrintFeatureFE 
Prints the value of one feature.

PrintObservationFE 
Prints the entire observation using MLObservation.toString()

RateFE 
Computes the derivative, i.e.

ResampleFE 
Creates new Observations by interpolating all features of the incoming Observations
whenever a feature (x) has increased by a specified amount (dx).

RunningTotalFE 
Compute the running total (i.e.

ScriptFE  
SegmentationFE 
A Segmentation splits one sequence of observations into several.

SetTagsFE 
Add tags to the Event corresponding to specified inputs.

SkipRepeatsFE 
Do not propagate observations if the previous and next observations have
the same value for the specified feature.

SlidingWindowFE 
This is a base class for smoothing feature extractors that need all the observations
for a specified amount of time before and after the current time before computing a result.

SlidingWindowFE.FractionMatching 
FractionMatching computes the fraction of inputs inside the time window matching the criteria.

SlidingWindowFE.FractionOfZeroes  
SlidingWindowFE.Min  
SmoothingFE 
Calculate the weighted average of past and future observations.

SpeedFE 
Computes the distance traveled per unit time, with no smoothing.

SplitGapsFE  
SplitGroupFE  
SplitGroupFE.LatLongMeters 
Two lat/long positions are near each other if the greatcircle distance between them is not more than some threshold distance in meters.

SplitGroupFE.ScalarDistance 
Two double vectors are near each other if the L2 norm (Euclidean distance) is not more than some threshold

StatisticFE 
This is a base class for implementing feature extractors that use one input, create one output, and only output a value on the final observation.

StatisticFE.BatchMedian  
StatisticFE.Count  
StatisticFE.Mean  
StatisticFE.Median  
StatisticFE.RootMeanSquare  
StatisticFE.Variance  
TagValueFE 
TagValueFE creates a feature from a tag value.

ThresholdFE  
TimeFE  
TimeFE.ObservationTime  
TimeFE.ProblemAge  
TrajectoryFE 
TrajectoryFE accrues the full history of another feature,
resamples it to a specified fixed length,
and then creates the output feature by concatenating all the past values together.

TurnRateFE 
Given a sequence of positions, computes the turn rate
(angular velocity) about the vertical axis.

Enum  Description 

AngleBetweenVectorsFE.Unit  
CogFoundryFE.Algorithm  
DistanceFE.DistanceType 
Exception  Description 

DTWSegmentationFE.QueryTooLongException 