Package 'otoclass' reference manual

Title:	Otolith Classification and Proportion Estimation
Description:	Methods for classification and analysis of otoliths along with methods for estimating species proportions in samples.
Authors:	Christoffer Moesgaard Albertsen [aut, cre]
Maintainer:	Christoffer Moesgaard Albertsen <[email protected]>
License:	GPL-2
Version:	0.8.0
Built:	2024-11-16 05:36:10 UTC
Source:	https://github.com/calbertsen/otoclass

Convert to otolith_contour_list

Description

Convert to otolith_contour_list

Usage

as.otolith_contour_list(x, ...)
as.otolith_contour_list(x, ...)

Arguments

`x`	otolith_contour object or output from read_image
`...`	other parameters

Value

otolith_contour_list object

Author(s)

Christoffer Moesgaard Albertsen

Assign unknown contours by binomial regression on elliptic fourier descriptors

Description

Assign unknown contours by binomial regression on elliptic fourier descriptors

Usage

assign_unknown(
  nimList,
  N = 60,
  delta = 0.01,
  lambda = 1,
  link = "probit",
  normalize_efd = FALSE
)
assign_unknown(
  nimList,
  N = 60,
  delta = 0.01,
  lambda = 1,
  link = "probit",
  normalize_efd = FALSE
)

Arguments

`nimList`	A list of normalized contours
`N`	Number of harmonics for the fourier descriptors
`delta`	passed to fcbf for feature selection
`lambda`	passed to fcbf for feature selection
`link`	Link function passed to `stats::glm`
`normalize_efd`	Should EFDs be normalized?

Value

A list of normalized contours where 'Unknown' Position attributes are replaced by 'Left' or 'Right'.

Author(s)

Christoffer Moesgaard Albertsen

Calculate Elliptical Fourier Descriptors

Description

Calculate Elliptical Fourier Descriptors

Usage

efd(x, N, ...)

## S3 method for class 'matrix'
efd(x, N, returnAsList = FALSE, normalize = FALSE, ...)

## S3 method for class 'otolith_contour'
efd(x, N, ...)

## S3 method for class 'otolith_image'
efd(x, N, ...)

## S3 method for class 'otolith_contour_list'
efd(x, N, ...)
efd(x, N, ...)

## S3 method for class 'matrix'
efd(x, N, returnAsList = FALSE, normalize = FALSE, ...)

## S3 method for class 'otolith_contour'
efd(x, N, ...)

## S3 method for class 'otolith_image'
efd(x, N, ...)

## S3 method for class 'otolith_contour_list'
efd(x, N, ...)

Arguments

`x`	contour object
`N`	Number of harmonics to calculate
`...`	other argiments
`returnAsList`	Should result be returned as a list that can be passed to `efd2coord`?
`normalize`	Normalize using method from Ferson et al. 1985?

Value

List of elliptical fourier descriptors

Author(s)

Christoffer Moesgaard Albertsen

References

Kuhl, F. P., Giardina, C. R. (1981) Elliptic Fourier Features of a Closed Contour. Computer Graphics and Image Processing. 18:236-258 Ferson, S., Rohlf, F. J., Koehn, R. K. (1985) Measuring Shape Variation of Two-Dimensional Outlines. Systematic Zoology. 34(1):59-68

Examples

Ellipsis <- function(theta,r1=1,r2=2){
  xx <- seq(0,2*pi,len=100)
  X <- cbind(r1*cos(xx),r2*sin(xx))
  R <- matrix(c(cos(theta),sin(theta),-sin(theta),cos(theta)),2,2)
  Y <- t(apply(X,1,function(x) as.vector(R %*% x)))
  return(Y)
}
efd(Ellipsis(pi/2),1)

efd(Ellipsis(pi/2),1,normalize = TRUE)

efd(Ellipsis(-pi/2),1,normalize = TRUE)

Ellipsis <- function(theta,r1=1,r2=2){
  xx <- seq(0,2*pi,len=100)
  X <- cbind(r1*cos(xx),r2*sin(xx))
  R <- matrix(c(cos(theta),sin(theta),-sin(theta),cos(theta)),2,2)
  Y <- t(apply(X,1,function(x) as.vector(R %*% x)))
  return(Y)
}
efd(Ellipsis(pi/2),1)

efd(Ellipsis(pi/2),1,normalize = TRUE)

efd(Ellipsis(-pi/2),1,normalize = TRUE)

Calculate XY-coordinates from Elliptical Fourier Descriptors

Description

Calculate XY-coordinates from Elliptical Fourier Descriptors

Usage

efd2coord(n, A, B, C, D, A0 = 0, C0 = 0)
efd2coord(n, A, B, C, D, A0 = 0, C0 = 0)

Arguments

`n`	Number of coordinates to calculate
`A`	A EFD coefficients
`B`	B EFD coefficients
`C`	C EFD coefficients
`D`	D EFD coefficients
`A0`	A0 EFD coefficients
`C0`	C0 EFD coefficients

Value

Matrix of coordinates

Author(s)

Christoffer Moesgaard Albertsen

Calculate XY-coordinates from Elliptical Fourier Descriptors

Description

Calculate XY-coordinates from Elliptical Fourier Descriptors

Usage

efd2coordOld(n, A, B, C, D, A0 = 0, C0 = 0)
efd2coordOld(n, A, B, C, D, A0 = 0, C0 = 0)

Arguments

`n`	Number of coordinates to calculate
`A`	A EFD coefficients
`B`	B EFD coefficients
`C`	C EFD coefficients
`D`	D EFD coefficients
`A0`	A0 EFD coefficients
`C0`	C0 EFD coefficients

Value

Matrix of coordinates

Author(s)

Christoffer Moesgaard Albertsen

Fast Correlation Based Filter for Feature Selection in High-Dimensional Data

Description

Fast Correlation Based Filter for Feature Selection in High-Dimensional Data

Usage

fcbf(
  train,
  group,
  delta,
  lambda = 1,
  nCuts = function(x) {     round(length(unique(x))/5) }
)
fcbf(
  train,
  group,
  delta,
  lambda = 1,
  nCuts = function(x) {     round(length(unique(x))/5) }
)

Arguments

`train`	A matrix of training data
`group`	A vector of groups
`delta`	Tuning parameter
`lambda`	Tuning parameter
`nCuts`	Function that calculates number of bins used for calculating correlations

Value

A vector of selected features

Author(s)

Christoffer Moesgaard Albertsen

References

http://pdf.aminer.org/000/335/746/feature_selection_for_high_dimensional_data_a_fast_correlation_based.pdf

Flip an otolith

Description

Flip an otolith

Usage

flip(x, axis, ...)
flip(x, axis, ...)

Arguments

`x`	otolith object
`axis`	axis to flip around
`...`	not used

Value

a flipped otolith object

Author(s)

Christoffer Moesgaard Albertsen

Fourier Series Basis for Splines

Description

Fourier Series Basis for Splines

Usage

fs(x, N, Boundary = range(x))
fs(x, N, Boundary = range(x))

Arguments

`x`	the predictor variable
`N`	Half the number of basis functions. Both a cos and a sin basis function will be added.
`Boundary`	Boundary points that determines a period

Value

A matrix of dimension 'c(length(x), 2*N)'.

Author(s)

Christoffer Moesgaard Albertsen

Prepare allele matrix for PCA analysis

Description

Prepare allele matrix for PCA analysis

Usage

gen2PCA(x, alleleMeans)
gen2PCA(x, alleleMeans)

Arguments

`x`	output from \linkread.gen
`alleleMeans`	Mean allele frequencies to impute. Is calculated if missing.

Value

allele data for PCA analysis

Author(s)

Christoffer Moesgaard Albertsen

Calculate group means for a data frame of covariates

Description

Calculate group means for a data frame of covariates

Usage

getGroupMeans(f, data, ...)

## S3 method for class 'mlld'
getGroupMeans(f, data, keep.cov = FALSE, ...)
getGroupMeans(f, data, ...)

## S3 method for class 'mlld'
getGroupMeans(f, data, keep.cov = FALSE, ...)

Arguments

`f`
`data`

Author(s)

Christoffer Moesgaard Albertsen

Calculate group proportions for a data frame of covariates

Description

Calculate group proportions for a data frame of covariates

Usage

getGroupProportion(f, data, ...)

## S3 method for class 'mlld'
getGroupProportion(f, data, randEff = TRUE, ...)
getGroupProportion(f, data, ...)

## S3 method for class 'mlld'
getGroupProportion(f, data, randEff = TRUE, ...)

Arguments

`f`
`data`

Author(s)

Christoffer Moesgaard Albertsen

Get a pixel matrix from an image file

Description

Get a pixel matrix from an image file

Usage

getPixelMatrix(file, grey = TRUE)
getPixelMatrix(file, grey = TRUE)

Arguments

`file`	Path to image file
`grey`	Should output be greyscale?

Value

A matrix of pixel values (0-255)

Author(s)

Christoffer Moesgaard Albertsen

Interpolate otolith coordinates through EFDs

Description

Interpolate otolith coordinates through EFDs

Usage

interpolate(x, N, Nefd, ...)
interpolate(x, N, Nefd, ...)

Arguments

`x`	otolith object
`N`	Number of coordinates to output
`Nefd`	Number of EFDs to use
`...`	not used

Value

an interpolated otolith object

Author(s)

Christoffer Moesgaard Albertsen

k Nearest Neighbour

Description

knn returns the votes to each level factor(group) from the kn nearest neighbours to test from train.

Usage

knn(train, group, test, kn = 1, dist = c("L2", "L1", "Linf"))
knn(train, group, test, kn = 1, dist = c("L2", "L1", "Linf"))

Arguments

`train`	A matrix of (continuous) features for the training sample
`group`	A vector of group labels for the training sample
`test`	A matrix of (continuous) features for the test sample
`kn`	Number of neighbours to consider
`dist`	Distance measure to use ("L1", "L2","Linf"). Default is Euclidian distance ("L2").

Maximum Likelihood Contour Discrimination

Description

Maximum Likelihood Contour Discrimination

Usage

mlcd(
  train,
  group,
  test,
  prior = as.vector(table(group))/length(group),
  penalty = 0,
  lambda = 0.4,
  Nefd = 60,
  correlatedCoordinates = FALSE,
  equalVariance = TRUE,
  silent = FALSE,
  control = list(iter.max = 1e+05, eval.max = 1e+05),
  ...
)
mlcd(
  train,
  group,
  test,
  prior = as.vector(table(group))/length(group),
  penalty = 0,
  lambda = 0.4,
  Nefd = 60,
  correlatedCoordinates = FALSE,
  equalVariance = TRUE,
  silent = FALSE,
  control = list(iter.max = 1e+05, eval.max = 1e+05),
  ...
)

Arguments

`train`	3D Array of training data (2 x number of contour points x number of observations)
`group`	Factor of training groups
`test`	3D array of test data
`prior`	Prior probability of groups
`penalty`	p to use for Lp penalty. Zero is no penalty.
`lambda`	Positive scalar factor for Lp penalty. Zero is no penalty.
`Nefd`	Number of EFDs to use for mean otolith
`correlatedCoordinates`	Should coordinates be correlated?
`equalVariance`	Should the variance be equal in the two coordinates?
`silent`	Should the TMB object be silent?
`control`	control parameters passes to nlminb
`...`	Other parameters

Value

a list of the result

Author(s)

Christoffer Moesgaard Albertsen

Maximum likelihood linear discrimination

Description

Maximum likelihood linear discrimination

Usage

mlld(
  y,
  group,
  data = NULL,
  formula = ~1,
  formulaCommon = ~1,
  formulaProportion = ~1,
  formulaLogScale = ~-1,
  equalLogScale = FALSE,
  lp_penalty = NA,
  lambda = 0.4,
  estimateLambda = TRUE,
  REML = FALSE,
  tMixture = 0,
  tDf = 5,
  estimateTMix = FALSE,
  sameTMix = FALSE,
  estimateTDf = FALSE,
  sameTDf = FALSE,
  independent = FALSE,
  silent = FALSE,
  control = list(iter.max = 1e+05, eval.max = 1e+05),
  drop.unused.levels = TRUE,
  onlyObj = FALSE,
  doSdreport = TRUE,
  getReportCovariance = TRUE,
  equalVariance = TRUE,
  confusionMatrixList = NULL,
  groupConversionList = list(),
  confusionLevelTypes = rep("Known", ifelse(is.null(ncol(group)), 1, ncol(group))),
  observationType = c("MVMIX", "SNP1", "SNP2", "SPL_AR1", "FS1_AR1", "SPL_CATEGORICAL"),
  forceMeanIncrease = FALSE,
  fixZeroGradient = TRUE,
  SNP2dm = FALSE,
  sameSNP2dm = TRUE,
  SPLknots = NULL,
  Nefd = 5,
  lower = list(),
  upper = list(),
  guessStartingValues = TRUE,
  imputeMissing = FALSE,
  ...
)
mlld(
  y,
  group,
  data = NULL,
  formula = ~1,
  formulaCommon = ~1,
  formulaProportion = ~1,
  formulaLogScale = ~-1,
  equalLogScale = FALSE,
  lp_penalty = NA,
  lambda = 0.4,
  estimateLambda = TRUE,
  REML = FALSE,
  tMixture = 0,
  tDf = 5,
  estimateTMix = FALSE,
  sameTMix = FALSE,
  estimateTDf = FALSE,
  sameTDf = FALSE,
  independent = FALSE,
  silent = FALSE,
  control = list(iter.max = 1e+05, eval.max = 1e+05),
  drop.unused.levels = TRUE,
  onlyObj = FALSE,
  doSdreport = TRUE,
  getReportCovariance = TRUE,
  equalVariance = TRUE,
  confusionMatrixList = NULL,
  groupConversionList = list(),
  confusionLevelTypes = rep("Known", ifelse(is.null(ncol(group)), 1, ncol(group))),
  observationType = c("MVMIX", "SNP1", "SNP2", "SPL_AR1", "FS1_AR1", "SPL_CATEGORICAL"),
  forceMeanIncrease = FALSE,
  fixZeroGradient = TRUE,
  SNP2dm = FALSE,
  sameSNP2dm = TRUE,
  SPLknots = NULL,
  Nefd = 5,
  lower = list(),
  upper = list(),
  guessStartingValues = TRUE,
  imputeMissing = FALSE,
  ...
)

Arguments

`...`
`proportionGroup`
`confusionGroup`

Author(s)

Christoffer Moesgaard Albertsen

Normalize an Otolith Image

Description

Normalize an Otolith Image

Usage

normalize_image(
  dat,
  n,
  datCompare = NULL,
  forceFlip = FALSE,
  flipByPosition = c("No", "Left", "Right", startByPolar = FALSE)
)
normalize_image(
  dat,
  n,
  datCompare = NULL,
  forceFlip = FALSE,
  flipByPosition = c("No", "Left", "Right", startByPolar = FALSE)
)

Arguments

`dat`	otolith_image object
`n`	Number of coordinates to output
`datCompare`	Data to compare with
`forceFlip`	Force a flip of the otolith?
`flipByPosition`	Position (Left/Right) to flip. Use 'No' if otoliths should not be flipped by position.

Value

A normalized Otolith Image

Author(s)

Christoffer Moesgaard Albertsen

Plot otolith contour

Description

Plot otolith contour

Usage

## S3 method for class 'otolith_contour'
plot(x, asUsed = FALSE, ...)
## S3 method for class 'otolith_contour'
plot(x, asUsed = FALSE, ...)

Arguments

`x`	Otolith contour
`asUsed`	Plot image as read; i.e. transformed
`...`	further arguments

Value

nothing

Author(s)

Christoffer Moesgaard Albertsen

Plot otolith contour list

Description

Plot otolith contour list

Usage

## S3 method for class 'otolith_contour_list'
plot(x, asUsed = FALSE, ask = TRUE, ...)
## S3 method for class 'otolith_contour_list'
plot(x, asUsed = FALSE, ask = TRUE, ...)

Arguments

`x`	Otolith contour list
`asUsed`	Plot image as read; i.e. transformed
`ask`	Ask before plotting next contour?
`...`	further arguments

Value

nothing

Author(s)

Christoffer Moesgaard Albertsen

Plot otolith image

Description

Plot otolith image

Usage

## S3 method for class 'otolith_image'
plot(x, asUsed = FALSE, ...)
## S3 method for class 'otolith_image'
plot(x, asUsed = FALSE, ...)

Arguments

`x`	Otolith image
`asUsed`	Plot image as read; i.e. transformed
`...`	further arguments

Value

nothing

Author(s)

Christoffer Moesgaard Albertsen

Signed area of a closed polygon

Description

The function uses the shoelace formula to calculate the signed area of a closed non-intersecting polygon.

Usage

polygon_area(X)
polygon_area(X)

Arguments

`X`	A numeric matrix with two columns of vertices

Value

Polygon area in same units as input

Author(s)

Christoffer Moesgaard Albertsen

Print otolith image contour

Description

Print otolith image contour

Usage

## S3 method for class 'otolith_contour'
print(x, ...)
## S3 method for class 'otolith_contour'
print(x, ...)

Arguments

`x`	Otolith image contour
`...`	further arguments

Value

nothing

Author(s)

Christoffer Moesgaard Albertsen

Print otolith image

Description

Print otolith image

Usage

## S3 method for class 'otolith_image'
print(x, ...)
## S3 method for class 'otolith_image'
print(x, ...)

Arguments

`x`	Otolith image
`...`	further arguments

Value

nothing

Author(s)

Christoffer Moesgaard Albertsen

LDA-like projection

Description

LDA-like projection

Usage

projection(x, ...)
projection(x, ...)

Arguments

`x`
`...`

Author(s)

Christoffer Moesgaard Albertsen

Read Otolith Images and Extract Contours

Description

Read Otolith Images and Extract Contours

Usage

read_image(
  file,
  noiseFactor = NULL,
  onlyOne = FALSE,
  minPixelDiff = 0.05 * min(nc, nr),
  extreme = TRUE,
  floodFillTolerance = 0.1,
  floodFillCol = 0,
  whiteBalance = 0.4,
  histogramAlpha = 0.4,
  borderBasedCutOff = FALSE,
  transformations = character(0),
  logisticTransformLocation = c("mean", "median", "borderMean", "borderMedian"),
  logisticTransformScale = 1,
  gaussianBlurSize = round(0.02 * min(nc, nr)),
  gaussianBlurSigma = gaussianBlurSize/10,
  pixelwise = FALSE,
  assignSinglesByPosition = TRUE,
  forceCenter = TRUE,
  minCountScale = 0,
  minCountForMax = 1e-04,
  reduceCutOffPercent = 0,
  zeroCutOffPercent = 0.05
)
read_image(
  file,
  noiseFactor = NULL,
  onlyOne = FALSE,
  minPixelDiff = 0.05 * min(nc, nr),
  extreme = TRUE,
  floodFillTolerance = 0.1,
  floodFillCol = 0,
  whiteBalance = 0.4,
  histogramAlpha = 0.4,
  borderBasedCutOff = FALSE,
  transformations = character(0),
  logisticTransformLocation = c("mean", "median", "borderMean", "borderMedian"),
  logisticTransformScale = 1,
  gaussianBlurSize = round(0.02 * min(nc, nr)),
  gaussianBlurSigma = gaussianBlurSize/10,
  pixelwise = FALSE,
  assignSinglesByPosition = TRUE,
  forceCenter = TRUE,
  minCountScale = 0,
  minCountForMax = 1e-04,
  reduceCutOffPercent = 0,
  zeroCutOffPercent = 0.05
)

Arguments

`file`	Image file path
`noiseFactor`	Scalar value determining cutoff value for background noise on image. If NULL, a value is calculated automatically.
`onlyOne`	Boolean value. Is there only one otolith on the image?
`minPixelDiff`	Minimum pixel difference between otoliths
`extreme`	Boolean value. Should pixel values be converted to 0/1?
`borderBasedCutOff`	Use a border based cut off?
`transformations`	Vector of image transformations: logistic, gaussianBlur, unsharp, floodFill
`logisticTransformLocation`	Location parameter for logistic transformation
`logisticTransformScale`	Scale parameter for logistic transformation
`gaussianBlurSize`	Size (n x n) for Gaussian blur kernel
`gaussianBlurSigma`	Variance parameter of Gaussian distribution used to blur
`pixelwise`	Boolean value. If TRUE, a pixel-wise algorithm is used to extract contours; otherwise, `grDevices::contourLines` is used.
`assignSinglesByPosition`	Should single otoliths be assigned to Left/Right based on position on image?
`minCountScale`	See details
`minCountForMax`	See details
`reduceCutOffPercent`	See details
`zeroCutOffPercent`	See details

Value

otolith image object

Author(s)

Christoffer Moesgaard Albertsen

Extract otolith contours from images

Description

Extract otolith contours from images

Usage

read_images(files, ...)
read_images(files, ...)

Arguments

`files`	file paths to read
`...`	Arguments passed to read_image

Value

an otolith_contour_list object

Author(s)

Christoffer Moesgaard Albertsen

Read genepop data files

Description

Read genepop data files

Usage

read.gen(
  f,
  pop.names,
  sort.loci = FALSE,
  sort.individuals = FALSE,
  NAlleleKeep = NA
)
read.gen(
  f,
  pop.names,
  sort.loci = FALSE,
  sort.individuals = FALSE,
  NAlleleKeep = NA
)

Arguments

`f`	file name
`pop.names`	population names. If missing, the ID of the last individual is used
`sort.loci`	Should loci be sorted by names?
`sort.individuals`	Should individuals be sorted by id?

Value

an allele array

Author(s)

Christoffer Moesgaard Albertsen

Rotate an otolith

Description

Rotate an otolith

Usage

rotate(x, angle, degrees = TRUE, ...)
rotate(x, angle, degrees = TRUE, ...)

Arguments

`x`	object to rotate
`angle`	Angles to rotate the object
`degrees`	Is the angle in degrees?
`...`	not used

Value

A rotated element

Author(s)

Christoffer Moesgaard Albertsen

Run shiny example included in the package

Description

Run shiny example included in the package

Usage

runShinyExample(name)
runShinyExample(name)

Arguments

name

Name of the shiny example (biascorrect, or proportion)

Value

Nothing

Author(s)

Christoffer Moesgaard Albertsen

Package 'otoclass'

Help Index

Convert to otolith_contour_list

Description

Usage

Arguments

Value

Author(s)

Assign unknown contours by binomial regression on elliptic fourier descriptors

Description

Usage

Arguments

Value

Author(s)

Calculate Elliptical Fourier Descriptors

Description

Usage

Arguments

Value

Author(s)

References

Examples

Calculate XY-coordinates from Elliptical Fourier Descriptors

Description

Usage

Arguments

Value

Author(s)

Calculate XY-coordinates from Elliptical Fourier Descriptors

Description

Usage

Arguments

Value

Author(s)

Fast Correlation Based Filter for Feature Selection in High-Dimensional Data

Description

Usage

Arguments

Value

Author(s)

References

Flip an otolith

Description

Usage

Arguments

Value

Author(s)

Fourier Series Basis for Splines

Description

Usage

Arguments

Value

Author(s)

Prepare allele matrix for PCA analysis

Description

Usage

Arguments

Value

Author(s)

Calculate group means for a data frame of covariates

Description

Usage

Arguments

Author(s)

Calculate group proportions for a data frame of covariates

Description

Usage

Arguments

Author(s)

Get a pixel matrix from an image file

Description

Usage

Arguments

Value

Author(s)

Interpolate otolith coordinates through EFDs

Description

Usage

Arguments

Value