false_detections: False detection filter

In jsta/glatos: A package for the Great Lakes Acoustic Telemetry Observation System

False detection filter

Description

Identify possible false detections based on "short interval" criteria (e.g., GLATOS 'min_lag') .

Usage

false_detections(det, tf, min_lag_col = "min_lag", show_plot = FALSE,
  ...)

Arguments

det	A glatos_detections object (e.g., produced by read_glatos_detections). OR: A data frame with one column containing 'min_lag' which for each detection record, is the smallest time (in seconds) to the next closest detection (either previous or subsequent) of the same transmitter on the same receiver. The name of the column containing 'min_lag' can be specified via min_lag_col; see below). OR (if min_lag is missing) A data farme containing detection data with the four columns described below. In that case, min_lag will be calculated using min_lag). detection_timestamp_utc Detection timestamps; MUST be of class POSIXct. transmitter_codespace A character string with transmitter code space (e.g., "A69-1061" for Vemco PPM coding"). transmitter_id A character string with transmitter ID code (e.g., "1363" for Vemco PPM coding"). receiver_sn A character vector with unique receiver serial number.
tf	A number indicating the time threshold (in seconds; e.g., Pincock's (2012) "short interval") for identifying possible false detections.
min_lag_col	A character string containing the name of the column in det that contains 'min_lag'.
show_plot	Indicates if a plot should be displayed showing the proportion of detections that exceed min_lag from min_lag = 1 to min_lag = 5 * tf.
...	Additional arguments passed to plot.

Details

Detections are identified as potentially false when min_lag > tf.

A new column (passed_filter), indicating if each record (row) passed the filter, is added to the input data frame.

This function was written specifically with GLATOS standard detection export in mind, but if min_lag is absent and min_lag_col is not specified, then min_lag will be calculated using min_lag.

A common rule of thumb for choosing tf for VEMCO PPM encoded transmitters is 30 times the nominal delay (e.g., 3600 s for a transmitter with a 120 s nominal delay) - see Pincock (2012).

When show_plot = TRUE then the plot may be used to assess sensitivity of the proportion of detections removed to the choice of tf.

Value

A data frame consisting of det with an additional column 'passed_filter' indicating if each detection did (1) or did not (0) pass the criteria.

Author(s)

T. R. Binder, edited by A. Dini

References

Pincock, D.G., 2012. False detections: what they are and how to remove them from detection data. Vemco Division, Amirix Systems Inc., Halifax, Nova Scotia.
http://www.vemco.com/pdf/false_detections.pdf

Simpfendorfer, C.A., Huveneers, C., Steckenreuter, A., Tattersall, K., Hoenner, X., Harcourt, R. and Heupel, M.R., 2015. Ghosts in the data: false detections in VEMCO pulse position modulation acoustic telemetry monitoring equipment. Animal Biotelemetry, 3(1), p.55.
https://animalbiotelemetry.biomedcentral.com/articles/10.1186/s40317-015-0094-z

See Also

min_lag

Examples

#get path to example detection file
det_file <- system.file("extdata", "walleye_detections.csv",
                         package = "glatos")
det <- read_glatos_detections(det_file)

det <- false_detections(det, 3600)
head(det)

#plot sensitivity to tf
det <- false_detections(det, 3600, show_plot = TRUE)

jsta/glatos documentation built on July 11, 2022, 7:01 a.m.