Type: Package
Title: Processing and Visualization of Video Plankton Recorder Data
Version: 0.2.3
Maintainer: Emily O'Grady <vprrcontact@gmail.com>
Description: An oceanographic data processing package for analyzing and visualizing Video Plankton Recorder data. This package was developed at 'Bedford Institute of Oceanography'. Functions are designed to process automated image classification output and create organized and easily portable data products.
License: MIT + file LICENSE
Encoding: UTF-8
LazyData: true
RoxygenNote: 7.1.2
Depends: R (≥ 2.10)
Imports: ggplot2, oce, dplyr, interp, magick, gsw, tidyr, stringr, metR, gridExtra, lubridate, cmocean, withr, usethis
Suggests: knitr, rmarkdown, testthat, markdown
VignetteBuilder: knitr
BuildVignettes: true
URL: https://eogrady21.github.io/vprr/
NeedsCompilation: no
Packaged: 2022-07-12 12:14:19 UTC; ChisholmE
Author: Emily O'Grady [aut, cre, cph], Kevin Sorochan dev [aut], Catherine Johnson [aut]
Repository: CRAN
Date/Publication: 2022-07-12 14:20:02 UTC

Get bin averages for VPR and CTD data

Description

Bins CTD data for an individual cast to avoid depth averaging across tow-yo's

Usage

bin_calculate(data, binSize = 1, imageVolume, rev = FALSE)

Arguments

data

ctd data frame object including scan, salinity, temperature, depth, conductivity, time, fluor_ref, turbidity_ref, turbidity_mv, altitude, cast_id, n_roi

binSize

the height of bins over which to average, default is 1 metre

imageVolume

the volume of VPR images used for calculating concentrations (mm^3)

rev

logical value, if TRUE, binning will begin at bottom of each cast, this controls data loss due to uneven binning over depth. If bins begin at bottom, small amounts of data may be lost at the surface of each cast, if binning begins at surface (rev = FALSE), small amounts of data may be lost at bottom of each cast

Details

Image volume calculations can change based on optical setting of VPR as well as autodeck setting used to process images For IML2018051 (S2) image volume was calculated as 108155 mm^3 by seascan (6.6 cubic inches) For COR2019002 S2 image volume was calculated as 83663 mm^3 and S3 image volume was calculated as 366082 mm^3. Used internally ( bin_cast ) after ctd_cast on a single ascending or descending section of VPR cast

Note

binSize should be carefully considered for best results

Depth is used for calculations! Please ensure depth is included in data frame using swDepth

Author(s)

E. Chisholm, K. Sorochan

Bin vpr data

Description

Formats oce style VPR data into depth averaged bins using ctd_cast and bin_calculate This function is used inside concentration_category

Usage

bin_cast(ctd_roi_oce, imageVolume, binSize, rev = FALSE)

Arguments

ctd_roi_oce

oce ctd format VPR data from vpr_oce_create

imageVolume

the volume of VPR images used for calculating concentrations (mm^3)

binSize

passed to bin_calculate, determines size of depth bins over which data is averaged

rev

logical value,passed to bin_calculate if TRUE, binning will begin at bottom of each cast, this controls data loss due to uneven binning over depth. If bins begin at bottom, small amounts of data may be lost at the surface of each cast, if binning begins at surface (rev = FALSE), small amounts of data may be lost at bottom of each cast

Details

Image volume calculations can change based on optical setting of VPR as well as autodeck setting used to process images For IML2018051 (S2) image volume was calculated as 108155 mm^3 by seascan (6.6 cubic inches) For COR2019002 S2 image volume was calculated as 83663 mm^3 and S3 image volume was calculated as 366082 mm^3

Value

A dataframe of depth averaged bins of VPR data over an entire cast with calculated concentration values

Binned concentrations

Description

This function produces depth binned concentrations for a specified taxa. Similar to bin_cast but calculates concentrations for only one taxa. Used inside vpr_roi_concentration

Usage

concentration_category(data, taxa, binSize, imageVolume, rev = FALSE)

Arguments

data

dataframe produced by processing internal to vpr_roi_concentration

taxa

name of taxa isolated

binSize

passed to bin_calculate, determines size of depth bins over which data is averaged

imageVolume

the volume of VPR images used for calculating concentrations (mm^3)

rev

Logical value defining direction of binning, FALSE - bins will be calculated from surface to bottom, TRUE- bins will be calculated bottom to surface

Details

Image volume calculations can change based on optical setting of VPR as well as autodeck setting used to process images For IML2018051 (S2) image volume was calculated as 108155 mm^3 by seascan (6.6 cubic inches) For COR2019002 S2 image volume was calculated as 83663 mm^3 and S3 image volume was calculated as 366082 mm^3

Author(s)

E. Chisholm

Isolate ascending or descending section of ctd cast

Description

This is an internal step required to bin data

Usage

ctd_cast(
  data,
  cast_direction = "ascending",
  data_type,
  cutoff = 0.1,
  breaks = NULL
)

Arguments

data

an oce ctd object

cast_direction

'ascending' or 'descending' depending on desired section

data_type

specify 'oce' or 'df' depending on class of desired output

cutoff

Argument passed to ctdFindProfiles

breaks

Argument passed to ctdFindProfiles

Value

Outputs either data frame or oce ctd object

Note

ctdFindProfiles arguments for minLength and cutOff were updated to prevent losing data (EC 2019/07/23)

Author(s)

K Sorochan, E Chisholm

VPR CTD data

Description

A dataframe including all CTD parameters from the VPR CTD, produced by vpr_ctd_read

Usage

ctd_dat_combine

Format

A dataframe with 15 variables

time_ms

Time stamp when ROI was collected (milliseconds)

conductivity

Conductivity collected by the VPR CTD

pressure

Pressure measured from the VPR CTD in decibars

temperature

Temperature measured from the VPR CTD in celsius

salinity

Salinity measured from the VPR CTD

fluor_ref

A reference fluorescence baseline provided in millivolts by the VPR CTD for calibrating fluorescence_mv data

fluorescence_mv

Fluorescence in millivolts from the VPR CTD (uncalibrated)

turbidity_ref

A reference turbidity baseline provided in millivolts for calibrating turbidity_mv

turbidity_mv

Turbidity in millivolts from the VPR CTD (uncalibrated)

altitude_NA

Altitude data from the VPR CTD

day

Day on which VPR data was collected (from AutoDeck)

hour

Hour during which VPR data was collected (from AutoDeck)

station

Station idnetifier provided during processing

sigmaT

Density caluclated from temperature, pressure and salinity data

depth

Depth in metres caluclated form pressure

Read CTD data (SBE49) and Fluorometer data from CTD- VPR package

Description

Internal use vpr_ctd_read

Usage

ctd_df_cols(x, col_list)

Arguments

x

full filename (ctd .dat file)

col_list

list of CTD data column names

Details

WARNING This is hard coded to accept a specific order of CTD data columns. The names and values in these columns can change based on the specific instrument and should be updated before processing data from a new VPR.

Text file format .dat file Outputs ctd dataframe with variables time_ms, conductivity, temperature, pressure, salinity, fluor_ref, fluorescence_mv, turbidity_ref, turbidity_mv, altitude_NA

Author(s)

K. Sorochan, E. Chisholm

VPR CTD data combined with tabulated ROIs

Description

A dataframe representing CTD data which has been merged with tabulated ROIs in each category, produced by vpr_ctdroi_merge

Usage

ctd_roi_merge

Format

A dataframe with 28 variables

time_ms

Time stamp when ROI was collected (milliseconds)

conductivity

Conductivity collected by the VPR CTD

pressure

Pressure measured from the VPR CTD in decibars

temperature

Temperature measured from the VPR CTD in celsius

salinity

Salinity measured from the VPR CTD

fluor_ref

A reference fluorescence baseline provided in millivolts by the VPR CTD for calibrating fluorescence_mv data

fluorescence_mv

Fluorescence in millivolts from the VPR CTD (uncalibrated)

turbidity_ref

A reference turbidity baseline provided in millivolts for calibrating turbidity_mv

turbidity_mv

Turbidity in millivolts from the VPR CTD (uncalibrated)

altitude_NA

Altitude data from the VPR CTD

day

Day on which VPR data was collected (from AutoDeck)

hour

Hour during which VPR data was collected (from AutoDeck)

station

Station identifier provided during processing

sigmaT

Density caluclated from temperature, pressure and salinity data

depth

Depth in metres caluclated form pressure

roi

ROI identification number

categories

For each category name (eg. bad_image_blurry, Calanus, krill), there is a line in the dataframe representing the number of ROIs identified in this category

n_roi_total

Total number of ROIs in all categories for each CTD data point

VPR data including CTD and ROI information

Description

An oce formatted CTD object with VPR CTD and ROI data from package example data set.

Usage

ctd_roi_oce

Format

An oce package format, a 'CTD' object with VPR CTD and ROI data (1000 data rows)

THIS FUNCTION HAS BEEN DEPRECATED

Description

pull roi measurements from all taxa, all files

Usage

getRoiMeasurements(taxafolder, nchar_folder, unit = "mm", opticalSetting)

Arguments

taxafolder

path to taxa folder (base – autoid folder)

nchar_folder

number of characters in basepath

unit

unit data will be output in, 'mm' (default – millimetres) or 'px' (pixels)

opticalSetting

VPR optical setting determining conversion between pixels and millimetres (options are 'S0', 'S1', 'S2', or 'S3')

Note

This function is very finicky, easily broken because it relies on character string splitting. taxaFolder argument should not end in a backslash, please check output carefully to ensure taxa names or ROI numbers have been properly sub string'd

INTERNAL USE ONLY quick data frame function from github to insert row inside dat frame

Description

INTERNAL USE ONLY quick data frame function from github to insert row inside dat frame

Usage

insertRow(existingDF, newrow, r)

Arguments

existingDF

data frame

newrow

new row of data

r

index of new row

Get vector to draw isopycnal lines on TS plot Used internally to create TS plots

Description

Get vector to draw isopycnal lines on TS plot Used internally to create TS plots

Usage

isopycnal_calculate(sal, pot.temp, reference.p = 0)

Arguments

sal

salinity vector

pot.temp

temperature vector in deg C

reference.p

reference pressure for calculation, set to 0

Note

: modified from source:https://github.com/Davidatlarge/ggTS/blob/master/ggTS_DK.R

Author(s)

E. Chisholm

Normalize a matrix

Description

take each element of matrix dived by column total

Usage

normalize_matrix(mat)

Arguments

mat

a matrix to normalize

Details

Make sure to remove total rows before using with VP data

Note

used internally for visualization of confusion matrices

Get conversion factor for pixels to mm for roi measurements

Description

Used internally

Usage

px_to_mm(x, opticalSetting)

Arguments

x

an aidmea data frame (standard) to be converted into mm from pixels

opticalSetting

the VPR setting determining the field of view and conversion factor between mm and pixels

Details

converts pixels to mm using conversion factor specific to optical setting

Options for opticalSetting are 'S0', 'S1', 'S2', or 'S3'

VPR ROI data

Description

A dataframe including VPR ROI data from the sample dataset, produced by vpr_autoid_read

Usage

roi_dat_combine

Format

A dataframe with 13 variables

roi

Unique ROI identifier - 8 digit

categories

For each category name (eg. bad_image_blurry, Calanus, krill), there is a line in the dataframe representing the number of ROIs identified in this category

time_ms

Time stamp when ROI was collected (milliseconds)

VPR measurement data calculated by Visual Plankton

Description

A data frame of measurement information for each ROI in the sample data set including long axis length, perimeter and area, produced by vpr_autoid_read

Usage

roimeas_dat_combine

Format

A data frame with 12 variables

roi

Unique ROI identifier - 10 digit

taxa

Category in which ROI has been classified by Visual Plankton

day_hour

day and hour in which data was collected (from Autodeck)

Perimeter

The perimeter of the ROI in millimeters

Area

The area of the ROI in millimeters

width1

Width at a first point of the ROI in millimetres (defined in more detail in VPR manual)

width2

Width at a second point of the ROI in millimetres (defined in more detail in VPR manual)

width3

Width at a third point of the ROI in millimetres (defined in more detail in VPR manual)

short_axis_length

The length in millimeters of the ROI along the shorter axis

long_axis_length

The length in millimeters of the ROI along the longer axis

station

Station identifier provided in processing

time_ms

Time stamp when ROI was collected in milliseconds

VPR size information dataframe

Description

A sample data frame of size information from Visual Plankton outputs, processed using vpr_ctdroisize_merge

Usage

size_df_f

Format

An object of class data.frame with 14 rows and 14 columns.

Details

@format A dataframe with 14 variables including

frame_ID

Unique identifier for each VPR frame

pressure

Pressure measured from the VPR CTD in decibars

temperature

Temperature measured from the VPR CTD in celsius

salinity

Salinity measured from the VPR CTD

sigmaT

Density calculated from temperature, salinity and pressure

fluorescence_mv

Fluorescence measured by the VPR CTD in millivolts (uncalibrated)

turbidity_mv

Turbidity measured by the VPR CTD in millivolts (uncalibrated)

roi

Unique ROI identification number - 10 digits, 8 digit millisecond time stamp and two unique digits to denote multiple ROIs within a millisecond

taxa

Category in which ROI has been classified by Visual Plankton

day_hour

Day and hour in which data was collected, from AutoDeck processing

long_axis_length

The length of the longest axis of the ROI image, measured by Visual Plankton

station

Station identifier provided during processing

time_ms

Time stamp when ROI was collected (milliseconds)

roi_ID

ROI identification number- 8 digit time stamp, without unique 2 digit ending

A binned data frame of concentration data per category

Description

A 'binned' dataframe from sample VPR data, including concentrations of each category, where each data point represents a 5 metre bin of averaged VPR data. Produced using vpr_roi_concentration

Usage

taxa_conc_n

Format

A dataframe with 21 variables

depth

Depth calculated from pressure in metres

min_depth

The minimum depth of the bin in metres

max_depth

The maximum depth of the bin in metres

depth_diff

The difference between minimum and maximum bin depth in metres

min_time_s

The minimum time in seconds of the bin

max_time_s

The maximum time in seconds of the bin

time_diff_s

The difference between minimum and maximum time in a bin, in seconds

n_roi_bin

The number of ROI observations in a bin

conc_m3

The concentration of ROIs in a bin, calculated based on image volume and number of frames per bin

temperature

Temperature measured from the VPR CTD in celsius (averaged within the bin)

salinity

Salinity measured from the VPR CTD (averaged within the bin)

density

sigma T density calculated from temperature, salinity and pressure (averaged within the bin)

fluorescence

Fluorescence measured by the VPR CTD in millivolts (uncalibrated) (averaged within the bin)

turbidity

Turbidity measured by the VPR CTD in millivolts (uncalibrated) (averaged within the bin)

avg_hr

The mean time in which bin data was collected, in hours

n_frames

The number of frames captured within a bin

vol_sampled_bin_m3

The volume of the bin sampled in metres cubed

toyo

Identifier of the tow-yo section which bin is a part of, either ascending or descending, appended by a number

max_cast_depth

The maximum depth of the entire VPR cast

taxa

The category in which ROIs in bin have been classified by Visual Plankton

station

Station identifier provided during processing

Plots normalized confusion matrix

Description

Plots normalized confusion matrix

Usage

vp_plot_matrix(cm, classes, type, addLabels = TRUE, threshold = NULL)

Arguments

cm

Confusion matrix (numeric)

classes

character list of classes present in confusion matrix (ordered)

type

character value 'NN', 'SVM' or 'Dual', appended to 'Confusion Matrix' to create title

addLabels

logical value whether to add percentage accuracy labels to plot (defaults to TRUE)

threshold

numeric value which determines the minimum value of frequency labelled on the plot on a normalized scale of 0-1 (useful for highlighting significant disagreement)

Value

a visualization of the confusion matrix, normalized

Author(s)

E. Chisholm

Function to visualize losses to unknown category due to disagreement in Dual classifier

Description

Makes confusion matrix like plot, where x axis represent SVM classification, y axis represent NN classification Allows visual summary of data lost to unknown category

Usage

vp_plot_unkn(cm, classes, threshold = 0, summary = TRUE, sample_size = NULL)

Arguments

cm

dual unknown confusion matrix from VP

classes

taxa groups in order, from VP

threshold

minimum value which will be labelled in plot

summary

logical to add text summary to plot E. Chisholm May 2019

sample_size

character string describes the sample size used to train the model being plotted (optional)

Checks manually created aid files for errors

Description

Removes any empty aid files after manual reclassification, checks for tow numbers and other metadata to match. Performs check to ensure measurement and ROI files are the same length

Usage

vpr_autoid_check(basepath, cruise, del)

Arguments

basepath

basepath to autoid folder eg. C:/data/CRUISENAME/autoid/

cruise

name of cruise which is being checked

del

Logical value, if TRUE, empty files will be deleted (see warning), if FALSE, files WILL NOT be deleted (they will be listed in output)

Details

WARNING: This function will delete empty aid and aidmeas files, permanently changing your directory. Consider making a back up copy before running this function.

Value

text file (saved in working directory) named CRUISENAME_aid_file_check.txt

Author(s)

E Chisholm

Copy VPR images into folders

Description

Organize VPR images into folders based on classifications provided by visual plankton

Usage

vpr_autoid_copy(basepath, day, hour, classifier_type, classifier_name, taxa)

Arguments

basepath

A file path to your autoid folder where VP data is stored eg. "C:\\data\\cruise_XXXXXXXXXXXXXXX\\autoid\\"

day

character string representing numeric day of interest

hour

character string representing hour of interest

classifier_type

character string representing the type of classifier (either 'svm', 'nn' or 'dual') from Visual Plankton

classifier_name

character string representing name of Visual Plankton classifier

taxa

optional list of character strings if you wish to only copy images from specific classification groups

Value

organized file directory where VPR images are contained with folders, organized by day, hour and classification, inside your basepath/autoid folder

Modifies aid and aid mea files based on manual reclassification

Description

Modifies aid and aid mea files based on manual reclassification

Usage

vpr_autoid_create(reclassify, misclassified, basepath, day, hour, mea = TRUE)

Arguments

reclassify

list of reclassify files (output from vpr_manual_classification())

misclassified

list misclassify files (output from vpr_manual_classification())

basepath

base path to auto ID folder eg 'E:/autoID_EC_07032019/'

day

day identifier for relevant aid & aidmeas files

hour

hour identifier for relevant aid & aidmeas files

mea

logical indicating whether or not there are accompanying measurement files to be created

examples

basepath <- 'E:/autoID_EC_07032019/' day <- '289' hr <- '08' day_hour_files <- paste0('d', day, '.h', hr) misclassified <- list.files(day_hour_files, pattern = 'misclassified_', full.names = TRUE) reclassify <- list.files(day_hour_files, pattern = 'reclassify_', full.names = TRUE) vpr_autoid_create(reclassify, misclassified, basepath)

Author(s)

E. Chisholm

Read VPR aid files

Description

Read aid text files containing ROI string information or measurement data and output as a dataframe

Usage

vpr_autoid_read(
  file_list_aid,
  file_list_aidmeas,
  export,
  station_of_interest,
  opticalSetting,
  warn = TRUE
)

Arguments

file_list_aid

a list object of aid text files, containing roi strings. Output from matlab Visual Plankton software.

file_list_aidmeas

a list object of aidmea text files, containing ROI measurements. Output from matlab Visual Plankton software.

export

a character string specifying which type of data to output, either 'aid' (roi strings) or 'aidmeas' (measurement data)

station_of_interest

Station information to be added to ROI data output, use NA if irrelevant

opticalSetting

Optional argument specifying VPR optical setting. If provided will be used to convert size data into mm from pixels, if missing size data will be output in pixels

warn

Logical, FALSE silences size data unit warnings

Details

Only outputs either ROI string information OR measurement data

Note

Full paths to each file should be specified

Author(s)

E. Chisholm & K. Sorochan

Examples


station_of_interest <- 'test'
dayhour <- c('d222.h03', 'd222.h04')

#' #VPR OPTICAL SETTING (S0, S1, S2 OR S3)
opticalSetting <- "S2"
imageVolume <- 83663 #mm^3

auto_id_folder <- system.file('extdata/COR2019002/autoid/', package = 'vprr', mustWork = TRUE)
auto_id_path <- list.files(paste0(auto_id_folder, "/"), full.names = TRUE)

#'   # Path to aid for each taxa
aid_path <- paste0(auto_id_path, '/aid/')
# Path to mea for each taxa
aidmea_path <- paste0(auto_id_path, '/aidmea/')

# AUTO ID FILES
aid_file_list <- list()
aidmea_file_list <- list()
for (i in 1:length(dayhour)) {
  aid_file_list[[i]] <-
    list.files(aid_path, pattern = dayhour[[i]], full.names = TRUE)
  # SIZE DATA FILES
  aidmea_file_list[[i]] <-
    list.files(aidmea_path, pattern = dayhour[[i]], full.names = TRUE)
}

aid_file_list_all <- unlist(aid_file_list)
aidmea_file_list_all <- unlist(aidmea_file_list)

 # ROIs
roi_dat_combine <-
  vpr_autoid_read(
    file_list_aid = aid_file_list_all,
    file_list_aidmeas = aidmea_file_list_all,
    export = 'aid',
    station_of_interest = station_of_interest,
    opticalSetting = opticalSetting,
    warn = FALSE
  )

# MEASUREMENTS
roimeas_dat_combine <-
  vpr_autoid_read(
    file_list_aid = aid_file_list_all,
    file_list_aidmeas = aidmea_file_list_all,
    export = 'aidmeas',
    station_of_interest = station_of_interest,
    opticalSetting = opticalSetting,
    warn = FALSE
 )

Get taxa ids from string

Description

Get taxa ids from string

Usage

vpr_category(x)

Arguments

x

A string specifying the directory of the "taxafolder", containing the taxa id

Value

A string of only the taxa id

Author(s)

K Sorochan

See Also

vpr_hour, vpr_day, vpr_roi

Examples

taxa_string <- 'C:/data/cruise/autoid/Calanus/d000/h00'
vpr_category(taxa_string)

Create a new taxa to be considered for classification after processing with VP

Description

creates empty directory structure to allow consideration of new taxa during vpr_manual_classification()

Usage

vpr_category_create(taxa, basepath)

Arguments

taxa

new taxa name to be added (can be a list of multiple taxa names)

basepath

basepath used for vpr_manual_classification

Value

empty directory structure using new taxa name inside basepath

Create a list of ctd files to be read

Description

Searches through typical VP directory structure

Usage

vpr_ctd_files(castdir, cruise, day_hour)

Arguments

castdir

root directory for ctd cast files

cruise

cruise name (exactly as in directory structure)

day_hour

vector of day-hour combinations (e.g, dXXX.hXX)

Details

Use with caution

Value

vector of ctd file paths matching days-hour combinations provided

Author(s)

E. Chisholm and K. Sorochan

Read and format CTD VPR data

Description

Acts as a wrapper for ctd_df_cols

Usage

vpr_ctd_read(ctd_files, station_of_interest, day, hour, col_list)

Arguments

ctd_files

full file paths to vpr ctd .dat files

station_of_interest

VPR station name

day

Day of interest, if not provided will be pulled from file path

hour

Hour of interest, if not provided will be pulled from file path

col_list

Optional list of CTD data column names

Details

Reads CTD data and adds day, hour, and station information. Calculates sigma T and depth variables from existing CTD data to supplement raw data. If there are multiple hours of CTD data, combines them into single dataframe.

WARNING ctd_df_cols is hard coded to accept a specific order of CTD data columns. The names and values in these columns can change based on the specific instrument and should be updated/confirmed before processing data from a new VPR.

Author(s)

E. Chisholm & K. Sorochan

Examples


station_of_interest <- 'test'

ctd_files <- system.file("extdata/COR2019002/rois/vpr5/d222", "h03ctd.dat",
package = "vprr", mustWork = TRUE)

ctd_dat_combine <- vpr_ctd_read(ctd_files, station_of_interest)

Add Year/ month/ day hour:minute:second information

Description

Calculate and record calendar dates for vpr data from day-of-year, hour, and time (in milliseconds) info. Will also add 'time_hr' parameter if not already present.

Usage

vpr_ctd_ymd(data, year, offset)

Arguments

data

VPR data frame from vpr_ctdroi_merge

year

Year of data collection

offset

time offset in hours between VPR CPU and processed data times (optional)

Value

a VPR data frame with complete date/time information in a new row named 'ymdhms'

Examples

year <- 2019
data('ctd_roi_merge')
dat <- vpr_ctd_ymd(ctd_roi_merge, year)

Merge CTD and ROI data from VPR

Description

Combines CTD data (time, hydrographic parameters), with ROI information (identification number) into single dataframe, aligning ROI identification numbers and taxa classifications with time and hydrographic parameters

Usage

vpr_ctdroi_merge(ctd_dat_combine, roi_dat_combine)

Arguments

ctd_dat_combine

a CTD dataframe from VPR processing from vpr_ctd_read

roi_dat_combine

a data frame of roi aid data from vpr_autoid_read

Author(s)

E. Chisholm & K. Sorochan

Examples

data('ctd_dat_combine')
data('roi_dat_combine')

ctd_roi_merge <- vpr_ctdroi_merge(ctd_dat_combine, roi_dat_combine)

Format CTD and Size data from VPR

Description

Format CTD and Meas data frames into combined data frame for analysis and plotting of size data

Usage

vpr_ctdroisize_merge(data, data_mea, taxa_of_interest)

Arguments

data

VPR dataframe from vpr_ctdroi_merge, with calculated variable sigmaT

data_mea

VPR size data frame from vpr_autoid_read

taxa_of_interest

a list of taxa of interest to be included in output dataframe

Value

A dataframe containing VPR CTD and size data

Examples


data("ctd_roi_merge")
data("roimeas_dat_combine")
category_of_interest = 'Calanus'

ctd_roi_merge$time_hr <- ctd_roi_merge$time_ms /3.6e+06

size_df_f <- vpr_ctdroisize_merge(ctd_roi_merge, data_mea = roimeas_dat_combine,
 taxa_of_interest = category_of_interest)

Get day identifier

Description

Get day identifier

Usage

vpr_day(x)

Arguments

x

A string specifying the directory and file name of the size file

Value

A string of only the day identifier (i.e., "dXXX")

Author(s)

K Sorochan

See Also

vpr_hour, vpr_roi, vpr_category

Examples

day_string <- 'C:/data/cruise/autoid/Calanus/d000/h00'
vpr_day(day_string)

Find day & hour info to match each station of interest for processing

Description

@author E. Chisholm and K. Sorochan

Usage

vpr_dayhour(stations, file)

Arguments

stations

a vector of character values naming stations of interest

file

CSV file containing 'day', 'hour', 'station', and 'day_hour' columns

Value

Vector of day-hour combinations corresponding to stations of interest

Get hour identifier

Description

Get hour identifier

Usage

vpr_hour(x)

Arguments

x

A string specifying the directory and file name of the size file

Value

A string of only the hour identifier (i.e., "hXX")

Author(s)

K Sorochan

See Also

vpr_day, vpr_roi, vpr_category

Examples

hour_string <- 'C:/data/cruise/autoid/Calanus/d000/h00'
vpr_hour(hour_string)

Explore images by depth and classification

Description

Pulls images from specific depth ranges in specific classification group

Usage

vpr_img_category(
  data,
  min.depth,
  max.depth,
  roiFolder,
  format = "list",
  taxa_of_interest
)

Arguments

data

data frame containing CTD and ROI data from vpr_ctdroi_merge, which also contains calculated variables sigmaT and time_hr

min.depth

minimum depth of ROIs you are interested in looking at

max.depth

maximum depth of ROIs you are interested in exploring

roiFolder

directory that ROIs are within (can be very general eg. C:/data, but will be quicker to process with more specific file path)

format

option of how images will be output, either as 'list' a list of file names or 'image' where images will be displayed

taxa_of_interest

character string of classification group from which to pull images

Remove ROI strings from aid and aidmeas files based on a manually organized folder of images

Description

Should be used after vpr_img_copy, and manual image removal from created folders

Usage

vpr_img_check(folder_dir, basepath)

Arguments

folder_dir

directory path to day hour folders containing manually reorganized images of a specific taxa eg. 'C:/data/cruise_IML2018051/krill/images/' where that folder contains '......d123.h01/' which contains manually sorted images of krill

basepath

directory path to original Visual Plankton files, specified down to the classification group. eg. 'C:/data/cruise_IML2018051/autoid/krill'

Image copying function for specific taxa of interest

Description

This function can be used to copy images from a particular taxa, day and hour into distinct folders within the auto id directory This is useful for visualizing the ROIs of a particular classification group or for performing manual tertiary checks to remove images not matching classification group descriptions.

Usage

vpr_img_copy(auto_id_folder, taxas.of.interest, day, hour)

Arguments

auto_id_folder

eg "D:/VP_data/IML2018051/autoid"

taxas.of.interest

eg. taxas.of.interest <- c('Calanus')

day

character, day of interest

hour

character, hour of interest

Explore VPR images by depth bin

Description

Allows user to pull VPR images from specific depth ranges, to investigate trends before classification of images into taxa groups

Usage

vpr_img_depth(data, min.depth, max.depth, roiFolder, format = "list")

Arguments

data

data frame containing CTD and ROI data from vpr_ctdroi_merge, which also contains calculated variables sigmaT and time_hr

min.depth

minimum depth of ROIs you are interested in looking at

max.depth

maximum depth of ROIs you are interested in exploring

roiFolder

directory that ROIs are within (can be very general eg. C:/data, but will be quicker to process with more specific file path)

format

option of how images will be output, either as 'list' a list of file names or 'image' where images will be displayed

Explore reclassified images

Description

Pull image from reclassified or misclassified files produced during vpr_manual_classification

Usage

vpr_img_reclassified(day, hour, base_dir, taxa_of_interest, image_dir)

Arguments

day

Character string, 3 digit day of interest of VPR data

hour

Character string, 2 digit hour of interest of VPR data

base_dir

directory path to folder containing day/hour folders in which misclassified and reclassified files are organized (eg.'C:/VPR_PROJECT/r_project_data_vis/classification files/') which would contain 'd123.h01/reclassified_krill.txt' )

taxa_of_interest

Classification group from which to pull images

image_dir

directory path to ROI images, eg. "E:\\data\\cruise_IML2018051\\", file separator MUST BE "\\" in order to be recognized

Value

folders of misclassified or reclassified images inside image_dir

Function to check results of classification manually

Description

Displays each image in day hour specified, prompts user to confirm or deny classification. If classification is denied, asks for a reclassification value based on available taxa

Usage

vpr_manual_classification(
  day,
  hour,
  basepath,
  taxa_of_interest,
  gr = TRUE,
  scale = "x300",
  opticalSetting = "S2",
  img_bright = TRUE
)

Arguments

day

day of interest in autoid

hour

hour of interest in autoid

basepath

file path to auto id folder eg 'E:/autoID_EC_07032019/'

taxa_of_interest

list of taxa folders you wish you sort through

gr

logical indicating whether pop up graphic menus are used (user preference - defaults to TRUE)

scale

argument passed to image_scale, default = 'x300'

opticalSetting

specifies optical setting of VPR, defining image frame size, current options are 'S0', 'S1', 'S2' (default), 'S3', see further info in details

img_bright

logical value indicating whether or not to include a blown out high brightness version of image (can be helpful for viewing dark field fine appendages)

Details

Optical Setting frame sizes: S0 = 7x7 mm, S1 = 14x14mm, S2 = 24x24mm, S3 = 48x48 mm. These settings define the conversion factor from pixels to millimetres and calculate image size for classification reference

Development

Create ctd oce object with vpr data

Description

Formats VPR data frame into oce format CTD object

Usage

vpr_oce_create(data)

Arguments

data

data frame of vpr data with variable names

'time_ms', 'fluorescence_mv', 'turbidity_mv', 'n_roi', 'sigmaT'

Author(s)

E. Chisholm

Examples

data('ctd_roi_merge')
oce_dat <- vpr_oce_create(ctd_roi_merge)

Make a balloon plot against a TS plot

Description

TS balloon plot with ROI concentration, sorted by taxa includes isopycnal line calculations

Usage

vpr_plot_TS(x, reference.p = 0, var)

Arguments

x

dataframe with temperature, salinity, number of rois (n_roi_bin)

reference.p

reference pressure (default at 0 for surface)- used to calculate isopycnals

var

variable on which size of points will be based, eg conc_m3 or n_roi_bin

Note

modified from source: https://github.com/Davidatlarge/ggTS/blob/master/ggTS_DK.R

Author(s)

E. Chisholm

Make a balloon plot

Description

Balloon plot against a TS plot with ROI concentration and sorted by taxa includes isopycnal line calculations. Version of vpr_plot_TS, with only relevant* taxa specified. *to current analysis and research objectives (See note).

Usage

vpr_plot_TScat(x, reference.p = 0)

Arguments

x

dataframe with temperature, salinity, number of rois named by taxa

reference.p

reference pressure (default at 0 for surface)- used to calculate isopycnals

Note

WARNING HARD CODED FOR 5 TAXA, CALANUS, KRILL, ECHINODERM LARVAE, SMALL COPEPOD, CHAETOGNATHS !! Uses isopycnal labelling method which does not label every contour

modified from source: https://github.com/Davidatlarge/ggTS/blob/master/ggTS_DK.R

Interpolated contour plot of particular variable

Description

Creates interpolated contour plot, can be used as a background for ROI or tow yo information

Usage

vpr_plot_contour(
  data,
  var,
  dup = "mean",
  method = "interp",
  labels = TRUE,
  bw = 1,
  cmo
)

Arguments

data

data frame needs to include time_hr, depth, and variable of choice (var)

var

variable in dataframe which will be interpolated and plotted

dup

if method == 'interp'. Method of handling duplicates in interpolation, passed to interp function (options: 'mean', 'strip', 'error')

method

Specifies interpolation method, options are 'interp' or 'oce', oce uses slightly different method (oce is least error prone)

labels

logical value indicating whether or not to plot contour labels

bw

bin width defining interval at which contours are labelled

cmo

name of a cmocean plotting theme, see ?cmocean for more information

Author(s)

E. Chisholm & Kevin Sorochan

Plot size frequency histogram

Description

Plot size frequency histogram

Usage

vpr_plot_histsize(data, param, title = NULL, bw = 0.1, xlim = NULL)

Arguments

data

size data from auto_measure_mm subset into taxas

param

size parameter of interest (corresponds to sub lists within data argument)

title

main title for plot, if left null will default based on parameter and taxa

bw

bin width, defines width of bars on histogram, defaults to 0.1, decrease for more detail

xlim

plot xlimit, defaults to min max of data if not provided

Note

param options are typically 'Perimeter', 'Area', 'width1','width2', 'width3', 'short_axis_length', 'long_axis_length'

Author(s)

E. Chisholm

Plots VPR profiles of temperature, salinity, density, fluorescence and concentration (by classification group)

Description

This plot allows a good overview of vertical distribution of individual classification groups along with reference to hydrographic parameters. Facet wrap is used to create distinct panels for each taxa provided

Usage

vpr_plot_profile(taxa_conc_n, taxa_to_plot, plot_conc)

Arguments

taxa_conc_n

A VPR data frame with hydrographic and concentration data separated by taxa (from vpr_roi_concentration)

taxa_to_plot

The specific classification groups which will be plotted, if NULL, will plot all taxa combined

plot_conc

Logical value whether or not to include a concentration plot (FALSE just shows CTD data)

Value

A gridded object of at least 3 ggplot objects

Size Frequency plots for VPR data

Description

This uses the hist plot function in base R to give a histogram of size (long axis length) frequency within a taxa. !!WARNING: this function uses hard coded plot attributes

Usage

vpr_plot_sizefreq(x, number_of_classes, colour_of_bar)

Arguments

x

a data frame with columns 'taxa', 'long_axis_length'

number_of_classes

numeric value passed to nclass argument in hist()

colour_of_bar

character value defining colour of plotted bars

Author(s)

K. Sorochan

Read prediction output from a CNN model

Description

Read prediction output from a CNN model

Usage

vpr_pred_read(filename)

Arguments

filename

model prediction output file (.txt) from vpr_transferlearn::save_output()

Value

a dataframe

Get roi ids from string

Description

Get roi ids from string

Usage

vpr_roi(x)

Arguments

x

A string specifying directory and file name of roi

Value

A string of only the 10 digit roi identifier

Author(s)

K Sorochan

See Also

vpr_hour, vpr_day, vpr_category

Examples


roi_string <- 'roi.0100000000.tif'
vpr_roi(roi_string)

Calculate VPR concentrations

Description

Calculates concentrations for each named taxa in dataframe

Usage

vpr_roi_concentration(
  data,
  taxas_list,
  station_of_interest,
  binSize,
  imageVolume
)

Arguments

data

a VPR dataframe as produced by vpr_ctdroi_merge

taxas_list

a list of character strings representing taxa present in the station being processed

station_of_interest

The station being processed

binSize

passed to bin_calculate, determines size of depth bins over which data is averaged

imageVolume

the volume of VPR images used for calculating concentrations (mm^3)

Examples


data('ctd_roi_merge')
ctd_roi_merge$time_hr <- ctd_roi_merge$time_ms /3.6e+06

taxas_list <- c('Calanus', 'krill')
binSize <- 5
station_of_interest <- 'test'
imageVolume <- 83663

taxa_conc_n <- vpr_roi_concentration(ctd_roi_merge, taxas_list,
station_of_interest, binSize, imageVolume)

Save VPR data as an as.oce object

Description

Save VPR data as an as.oce object

Usage

vpr_save(data, metadata)

Arguments

data

a VPR data frame

metadata

(optional) a named list of character values giving metadata values. If this argument is not provided user will be prompted for a few generic metadata requirements.

Details

This function will pass a VPR data frame object to an oce object. Using an oce object as the default export format for VPR data allows for metadata and data to be kept in the same, space efficient file, and avoid redundancy in the data frame. The function check for data parameters that may actually be metadata parameters (rows which have the same value repeated for every observation). These parameters will automatically be copied into the metadata slot of the oce object. The function will also prompt for a variety of required metadata fields. Depending on specific research / archiving requirements, these metadata parameters could be updated by providing the argument metadata.

Default metadata parameters include 'deploymentType', 'waterDepth', 'serialNumber', 'latitude', 'longitude', 'castDate', 'castStartTime', 'castEndTime', 'processedBy', 'opticalSetting', 'imageVolume', 'comment'.

Value

an oce CTD object with all VPR data as well as metadata

Examples

data("taxa_conc_n")
metadata <- c('deploymentType' = 'towyo', 'waterDepth' =
max(ctd_roi_merge$pressure), 'serialNumber' = NA, 'latitude' = 47,
'longitude' = -65, 'castDate' = '2019-08-11', 'castStartTime'= '00:00',
'castEndTime' = '01:00', 'processedBy' = 'E. Chisholm', 'opticalSetting' =
'S2', 'imageVolume' = 83663, 'comment' = 'test data')

oce_dat <- vpr_save(taxa_conc_n, metadata)
# save(oce_dat, file = vpr_save.RData') # save data

Bin VPR size data

Description

Calculates statistics for VPR measurement data in depth averaged bins for analysis and visualization

Usage

vpr_size_bin(data_all, bin_mea)

Arguments

data_all

a VPR CTD and measurement dataframe from vpr_ctdroisize_merge

bin_mea

Numerical value representing size of depth bins over which data will be combined, unit is metres, typical values range from 1 - 5

Value

a dataframe of binned VPR size data statistics including number of observations, median, interquartile ranges, salinity and pressure, useful for making boxplots

Examples


data('size_df_f')
vpr_size_bin(size_df_f, bin_mea = 5)

Data Summary Report

Description

Part of VP easy plot processing, prints data summary report to give quantitative, exploratory analysis of data

Usage

vpr_summary(all_dat, fn, tow = tow, day = day, hour = hour)

Arguments

all_dat

data frame containing VPR and CTD data including time_ms, time_hr, conductivity, temperature, pressure, salinity, fluorescence_mv, turbidity_mv, sigmaT

fn

file name to save data summary, if not provided, summary will print to console

tow

VPR tow number

day

julian day

hour

two digit hour (24 hr clock)

Author(s)

E Chisholm

Get size data from idsize files

Description

useful for getting size distribution of known rois from each taxa. gathers size information from idsize text files produced when training a new classifier in VP (Visual Plankton)

Usage

vpr_trrois_size(directory, taxa, opticalSetting)

Arguments

directory

cruise directory eg. 'C:/data/IML2018051/'

taxa

list of character elements containing taxa of interest

opticalSetting

VPR optical setting determining conversion between pixels and millimetres (options are 'S0', 'S1', 'S2', or 'S3')

mirror server hosted at Truenetwork, Russian Federation.