| export_banter {PAMpal} | R Documentation |
Export Data for a BANTER Model
Description
Exports data from an AcousticStudy into the format required to run a BANTER model from the "banter" package
Usage
export_banter(
x,
dropVars = NULL,
dropSpecies = NULL,
training = TRUE,
verbose = TRUE
)
Arguments
x |
a AcousticStudy object or a list of AcousticEvent objects |
dropVars |
a vector of the names of any variables to remove |
dropSpecies |
a vector of the names of any species to exclude |
training |
logical flag whether or not this will be used as a
training data set, or a value between 0 and 1 specifying what percent
of the data should be used for training (with the rest set aside
for testing). If TRUE or greater than 0, must contain species ID.
NOTE: if value is not 0, 1, |
verbose |
logical flag to show summary and informational messages |
Value
a list with three items, events, detectors, and
na. If value of training is not 0, 1, TRUE, or
FALSE, output will be split into training and
test lists that contain events and detectors.
events is a dataframe with two columns. event.id is a
unique identifier for each event, taken from the names of the event
list. species is the species classification, taken from the
species slot labelled id. detectors is a list
of data frames containing all the detections and measurements. There is
one list for each unique detector type found in the detectors slots
of x. The data frames will only have columns with class
numeric, integer, factor, or logical, and
will also have columns named UID, Id, parentID,
sampleRate, Channel, angle, and angleError,
removed so that these are not treated as parameters for the banter random
forest model. The dataframes will also have columns event.id and
call.id added. na contains the UIDs and Binary File names
for any detections that had NA values. These cannot be used in the
random forest model and are removed from the exported dataset.
Author(s)
Taiki Sakai taiki.sakai@noaa.gov
Examples
# setting up example data
data(exStudy)
exStudy <- setSpecies(exStudy, method='pamguard')
banterData <- export_banter(exStudy)
# drop some variables
banterLess <- export_banter(exStudy, dropVars = c('peak', 'duration'))