ecoscope.analysis.immobility#
Module Contents#
- class ecoscope.analysis.immobility.ImmobilityProfile[source]#
threshold_radius: radius of cluster in metres. Defaults to 13m. threshold_time: time in seconds the track is expected to be stationary. Defaults to 18000 seconds (5 hours) threshold_probability: the proportion of observations in a sample which must be inside a cluster to generate a CRITICAL. Default 0.8
- threshold_radius: int = 13#
- threshold_time: int = 18000#
- threshold_probability: float = 0.8#
- class ecoscope.analysis.immobility.Immobility[source]#
- classmethod calculate_immobility(immobility_profile, relocs)[source]#
Function to search for immobility within a movement trajectory. Assumes we start with a filtered trajectory spanning some period of time. The algorithm will work backwards through the trajectory’s relocations and build a cluster. Looks to see if the cluster characteristics match immobility criteria (ie., timespan is gte than the threshold_time, and the cluster probability is gte to the threshold_probability) Note that this is a simplified version of the full clustering algorithm since it’s only looking at data within threshold time and will not figure out the true start of an immobility without looking backwards through all possible points TODO: include more info about the immobility result: 1) immobility start time 2) immobility probability 3) immobility cluster fix count 4) algorithm provenance
- Parameters:
immobility_profile (ImmobilityProfile) – setting for immobility
relocs (ecoscope.base.Relocations) – set for fixes for given subject.
- Return type:
Dict