Category: Transform

CollapseTemporalGaps

Removes empty days by shifting data to fill those empty days.

Input name Type Description
startAt an instant; required Start date for all traces
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

DurationSplitting

Input name Type Description
duration a duration; required Maximum duration of each trace
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

EnforceDuration

Longer traces will be truncated, shorter traces will be discarded.

Input name Type Description
minDuration a duration; optional Minimum duration of a trace
maxDuration a duration; optional Maximum duration of a trace
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

EnforceSize

Larger traces will be truncated, smaller traces will be discarded.

Input name Type Description
minSize a 32-bit integer; optional Minimum number of events in each trace
maxSize a 32-bit integer; optional Maximum number of events in each trace
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

GaussianKernelSmoothing

Apply gaussian kernel smoothing on a trace, attenuating the impact of noisy observations.

Input name Type Description
omega a duration; required Bandwidth
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

ModuloSampling

It will ensure that the final number of events is exactly (+/- 1) the one required, and that events are regularly sampled (i.e., one out of x).

Input name Type Description
n a 32-bit integer; required Number of events to keep
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

PoisExtraction

Input name Type Description
diameter a distance; required Clustering maximum diameter
duration a duration; required Clustering minimum duration
minPoints a 32-bit integer; optional; default: 0 Minimum number of times a cluster should appear to consider it
data a file; required Input traces dataset
Output name Type Description
data a file Transformed dataset

SequentialSplitting

Input name Type Description
percentBegin a 32-bit integer; required Percentage of events at which a trace begins
percentEnd a 32-bit integer; required Percentage of events at which a trace ends
complement a boolean; optional; default: false Whether to take the complement trace
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

SizeSplitting

Input name Type Description
size a 32-bit integer; required Maximum number of events allowed in each trace
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

SpatialGapSplitting

Input name Type Description
distance a distance; required Maximum distance between two consecutive events
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

SpatialSampling

If the distance is less than a given threshold, records will be discarded until the next point that fulfills the minimum distance requirement.

Input name Type Description
distance a distance; required Minimum distance between two consecutive events
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

TemporalGapSplitting

Input name Type Description
duration a duration; required Maximum duration between two consecutive events
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

TemporalSampling

If the duration is less than a given threshold, events will be discarded until the next point that fulfills the minimum duration requirement.

Input name Type Description
duration a duration; required Minimum duration between two consecutive events
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset

UniformSampling

Perform a uniform sampling on traces, keeping each event with a given probability.

Input name Type Description
probability a 64-bit float; required Probability to keep each event
data a file; required Input dataset
Output name Type Description
data a file Transformed dataset