Extract information from deployment log
ss_read_log.Rd
Extract information from deployment log
Arguments
- path
File path to the Log folder.
- path_config
File path (including name and extension) of the water quality configuration table. Default is water_quality_configuration_table.xlsx in the tracking_sheets folder. Data must be stored in an xlsx file in a tab named "deployments". Columns must be Waterbody, Station_Name, and Depl_Date. One and only one row must match the waterbody, station, and deployment date recorded in the Log. Options for the configuration column are sub-surface buoy, surface buoy, attached to gear, attached to fixed structure, floating dock, unknown, and calval. Use of the "unknown" entry is discouraged. HINT: for optimal speed, make sure the table is NOT filtered in the excel file.
The
path_config
argument is ignored for new deployment logs that have an appropriate entry in theConfiguration
column.Configurations recorded as
NA
will be converted to "unknown" with aWarning
.When there is no
Configuration
in the log and the deployment is not found in the configuration table, the configuration will berecoreded as "unknown" with aWarning
.If more than one configuration is detected in the log or the configuration table, the function will stop with an
Error
.
Value
Returns a list with 4 elements. deployment_dates
is a data
frame with two columns: start_date
(the date of deployment) and
end_date
(date of retrieval). area_info
is a data frame with
five columns: county
, waterbody
, latitude
,
longitude
, station
, and lease
. sn_table
is a
data frame with three columns: log_sensor
(sensor name as recorded
in the log), sensor_serial_number
, and depth
(sensor depth
below the surface at low tide, from the Sensor_Depth column).
string_configuration
is a character string indicating how the sensor
string was moored (i.e., whether the sensors float with the tide).
Details
These columns are going to change with the new logs. Deprecate this function or save it as _old for backwards compatibility.
The log must be saved in a folder called Log in .csv, .xlsx or .xls format, and must include the following columns:
Deployment_Waterbody
: waterbody where string was deployed
Location_Description
: the station name
Lease#
: If located on an aquaculture site, the lease number (NA
otherwise)
Deployment
: deployment date, in the order "Ymd"
Retrieval
: retrieval date, in the order "Ymd"
Logger_Latitude
: The latitude at which the string was deployed
Logger_Longitude
The longitude at which the string was deployed (must
be a negative value)
Logger_Model
The type of sensor; see below for options
Serial#
The sensor serial number
Sensor_Depth
: Depth at which the sensor was deployed
All other columns will be ignored.
Entries in the Logger_Model
column must include the string
"aquameasure", "hobo", "tidbit", or "vr2ar" (not case sensitive).
The function will stop with an Error if there is more than one eligible file (csv, .xlsx or .xls) in the Log folder.
The function will stop with an Error if there if the Logger_Longitude
is a positive value.
A Warning message is printed to the console when the function does not recognize a sensor in the log.
A message is printed to the console when hobo, aquameasure, or vemco sensors are not found in the log.
A message is printed to the console if there is more than one unique entry in
Deployment_Waterbody
, Location_Description
, Deployment
,
Retrieval
, Logger_Latitude
, or Logger_Longitude
.
See also
Other compile:
ss_compile_aquameasure_data()
,
ss_compile_deployment_data()
,
ss_compile_hobo_data()
,
ss_compile_vemco_data()