Runs a number of quality assurance / quality control (QAQC) checks on a folder of recording files to identify potential problems. These include checking the start and end times of files for consistency to identify potential data gaps, measuring sound levels in each file to identify potential recorder issues, and (if applicable) checking battery and temperature data to identify potential instrument failure. Can also create spectrogram images throughout the deployment to aid in visually checking for problems or noise.
evaluateDeployment(
dir,
excludeDirs = c("Post_Retrieval_Data", "Pre_Deployment_Data"),
sampleWindow = c(60, 120),
channel = 1,
sensitivity = NA,
calibration = NULL,
timeRange = NULL,
name = NULL,
subDirPattern = NULL,
outDir = NULL,
nSpectrograms = 0,
specLength = 1800,
panelLength = 300,
log = FALSE,
progress = TRUE,
verbose = TRUE
)a dataframe of the QAQC metric outputs for each recording file
folder or folders containing recordings and optionally Soundtrap
.log.xml files. All .wav and .log.xml files within dir will be
analysed, as well as all files in each subfolder of dir (only
going down one level).
the names of any subfolders within dir that
should be excluded (e.g. if dir contains folders "Recordings" and
"Clips" then excludeDirs="Clips" would result in only the "Recordings"
folder being analysed
start and end (in seconds) of the time window to use
for analysis, e.g. c(40, 100) will use a 60 second window starting
40 seconds into the file
channel number of recording files to use for analysis
the sensitivity of the recording device in dB, this is typically a large negative number
if not NULL, the frequency dependent calibration
to apply. Must have "frequency" and "gain" (in dB), can either be a .tf
file, a CSV file with columns for frequency and gain, or a dataframe with
columns frequency and gain
if not NULL, a vector of two POSIXct times identifying
the expected start and end times of the deployment. If the actual start and
end times of of the recording files are earlier or later than these, then
a warning will be issued and no calculations will be done, returning
NULL
a name to assign for this deployment, used for plot labeling,
logging, and stored as projectName with the output dataframe. If
left as NULL, then the basename of dir will be used.
if not NULL, a pattern to use for selecting
which subfolders of dir to use for analysis. E.g. if folders
"Site1_Recordings" and "Site2_Recordings" both exist in dir, then
subDirPattern="^Site1" would result in analysing only the first
folder of recordings
if not NULL, a directory to store outputs. Outputs include
plots, a CSV of the calculated metrics, and a .txt log file if log=TRUE
number of spectrogram images to generate. Recording files
for creating the spectrograms will be approximately evenly spaced throughout
the deployment - if nSpectrograms=5, then the first recording will be
randomly chosen from the first 20% of files, the second from the second 20%,
etc.
length of spectrograms to create, in seconds
length of each panel of the spectrogram plot, in seconds.
This must be less than specLength, and is used to avoid creating
awkwardly long plots. If specLength=360 and panelLength=60,
then the result will be a 6 panel plot where each section is 60 seconds long
if TRUE and outDir is not NULL, then a text
file named "(name)_EvaluateRecorder_LogFile.txt" will be created in
outDir logging progress and warning messages
logical flag to show a progress bar
logical flag to show some messages