Package {specmine}

specmine: Metabolomics data analysis tools

Description

Tools for metabolomics and spectral data analysis and mining.

Value

No return value, called for side effects.

Author(s)

Maintainer: Pedro Fontão pedrofontao812004@gmail.com

Authors:

• Pedro Fontão pedrofontao812004@gmail.com

• Christopher Costa chrisbcl@hotmail.com

• Marcelo Maraschin mtocsy@gmail.com

• Miguel Rocha mrocha@di.uminho.pt

• Sara Cardoso saracardoso501@gmail.com

• Telma Afonso telma.afonso94@gmail.com

• Bruno Pereira pereirinha_bp@hotmail.com

Other contributors:

• C. Beleites [copyright holder]

• Jie Hao [copyright holder]

See Also

Useful links:

• https://github.com/PedroFontao/specmine

• Report bugs at https://github.com/PedroFontao/specmine/issues

Examples

## Not run: 
data(package = "specmine")

## End(Not run)

Aggregate samples

Description

Aggregate samples according to an aggregate function like mean, median, etc. This can be used to merge replicates.

Usage

aggregate_samples(dataset, indexes, aggreg.fn = "mean", meta.to.remove = c())

Arguments

Value

A dataset object in which samples have been aggregated according to 'indexes' and 'aggreg.fn', with updated 'data', 'metadata', and preserved dataset-level fields such as labels, type, and description.

Examples

data <- matrix(
  c(1, 2, 10, 12,
    3, 4, 14, 16),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3", "s4"))
)
metadata <- data.frame(
  class = factor(c("A", "A", "B", "B")),
  batch = c(1, 1, 2, 2),
  row.names = c("s1", "s2", "s3", "s4")
)
dataset <- list(
  data = data,
  metadata = metadata,
  labels = list(),
  type = "example",
  description = "toy dataset"
)
aggregate_samples(dataset, c(1, 1, 2, 2), "mean")

Analysis of variance

Description

Performs one-way ANOVA for all variables in a dataset.

Usage

aov_all_vars(
  dataset,
  column.class,
  doTukey = TRUE,
  write.file = FALSE,
  file.out = NULL
)

Arguments

Value

A data frame with ANOVA results for all variables.

Examples

datamatrix <- matrix(
  c(10, 11, 20, 21,
    5,  6,  5,  6),
  nrow = 2,
  byrow = TRUE,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3", "s4"))
)
metadata <- data.frame(
  group = factor(c("A", "A", "B", "B")),
  row.names = c("s1", "s2", "s3", "s4")
)
dataset <- list(data = datamatrix, metadata = metadata)
aov_all_vars(dataset, "group", doTukey = FALSE)

Analysis of variance for one variable

Description

Performs one-way ANOVA for a single variable in a dataset.

Usage

aov_one_var(dataset, x.val, groups, doTukey = TRUE)

Arguments

Value

A list with p-value, -log10(p), FDR, and optional Tukey results.

Examples

datamatrix <- matrix(
  c(10, 11, 20, 21,
    5,  6,  5,  6),
  nrow = 2,
  byrow = TRUE,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3", "s4"))
)
metadata <- data.frame(
  group = factor(c("A", "A", "B", "B")),
  row.names = c("s1", "s2", "s3", "s4")
)
dataset <- list(data = datamatrix, metadata = metadata)
aov_one_var(dataset, "x1", metadata$group, doTukey = FALSE)

Apply by group

Description

Applies a function to the variables of samples belonging to a given group.

Usage

apply_by_group(dataset, fn.to.apply, metadata.var, var.value)

Arguments

Value

A named vector or matrix, depending on the output of 'fn.to.apply', containing the values obtained by applying the function to each variable across the samples belonging to the selected group.

Examples

data <- matrix(
  c(1, 2, 3, 4, 5, 6),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  group = c("control", "control", "case"),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
apply_by_group(dataset, mean, "group", "control")

Apply by groups

Description

Applies a function to groups defined by a metadata variable.

Usage

apply_by_groups(
  dataset,
  metadata.var,
  fn.to.apply = "mean",
  variables = NULL,
  variable.bounds = NULL
)

Arguments

Value

A matrix-like object with one row per selected variable and one column per group defined by 'metadata.var'. Each cell contains the result of applying 'fn.to.apply' to the values of that variable within the corresponding group.

Examples

data <- matrix(
  c(1, 2, 3, 4, 5, 6),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  group = c("control", "control", "case"),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
apply_by_groups(dataset, "group", mean)

Auto-exported function: baseline_correction

Description

Auto-exported function: baseline_correction

Usage

baseline_correction(dataset, method = "modpolyfit", ...)

Arguments

Value

A dataset object containing the baseline-corrected data. The returned object has the same overall structure as the input dataset, with corrected intensity values stored in dataset$data, original row and column names preserved, and the description updated to record the baseline correction step.

Examples

datamatrix <- matrix(
  c(5, 6, 7,
    6, 7, 8,
    7, 8, 9,
    8, 9, 10,
    9, 8, 7,
    8, 7, 6,
    7, 6, 5,
    6, 5, 4),
  nrow = 8,
  byrow = TRUE,
  dimnames = list(as.character(1:8), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = c("A", "A", "B"),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(
  data = datamatrix,
  metadata = metadata,
  description = "toy spectra",
  labels = list(x = "ppm", val = "intensity")
)
corrected <- try(
  baseline_correction(dataset, method = "modpolyfit"),
  silent = TRUE
)
corrected

Get kegg codes from chebi codes:

Description

Auto-exported function: convert_chebi_to_kegg

Usage

convert_chebi_to_kegg(chebi_codes)

Arguments

Value

A named character vector with the KEGG compound identifiers corresponding to the input ChEBI codes. The vector values are KEGG compound codes prefixed with "cpd:", and the element names are the matched compound names.

Examples

## Not run: 
convert_chebi_to_kegg(c("CHEBI:15377"))

## End(Not run)

Auto-exported function: convert_from_chemospec

Description

Auto-exported function: convert_from_chemospec

Usage

convert_from_chemospec(csobj, type = "undefined", description = "")

Value

A dataset object converted from a ChemoSpec object, containing the transposed data matrix, metadata, sample names, and axis/unit labels.

Examples

csobj <- list(
  data = matrix(c(10, 20, 30, 40), nrow = 2, byrow = TRUE),
  freq = c(1, 2),
  groups = c("A", "B"),
  names = c("sample1", "sample2"),
  unit = c("ppm", "intensity")
)
dataset <- convert_from_chemospec(csobj)
class(dataset)

Get kegg codes from hmdb codes:

Description

Get kegg codes from hmdb codes:

Usage

convert_hmdb_to_kegg(hmdb_codes)

Arguments

Value

A named character vector with the KEGG compound identifiers corresponding to the input HMDB codes. The vector values are KEGG compound codes prefixed with "cpd:", and the element names are the matched compound names.

Examples

## Not run: 
convert_hmdb_to_kegg(c("HMDB0000122"))

## End(Not run)

Convert KEGGPathway object to graph object

Description

Convert KEGGPathway object to graph object

Usage

convert_keggpathway_2_reactiongraph(pathObj)

Arguments

Value

A graph object representing the reactions in the input KEGGPathway object. The returned object is suitable for graph-based inspection or for use in downstream visualization functions.

Examples

## Not run: 
path <- get_MetabolitePath("hsa00010")
graph <- convert_keggpathway_2_reactiongraph(path)
class(graph)

## End(Not run)

Get kegg codes from spcmnm codes:

Description

Get kegg codes from spcmnm codes:

Usage

convert_multiple_spcmnm_to_kegg(spcmnm_codes)

Arguments

Value

A named character vector with the KEGG compound identifiers corresponding to the input SPCMNM codes. The vector values are KEGG compound codes prefixed with "cpd:", and the element names are the matched compound names.

Examples

## Not run: 
convert_multiple_spcmnm_to_kegg(c("SPCM00001"))

## End(Not run)

Count missing values

Description

Returns the total number of missing values in the dataset.

Usage

count_missing_values(dataset)

Arguments

Value

A single numeric value giving the total count of 'NA' entries present in 'dataset$data'.

Examples

data <- matrix(
  c(1, NA, 3, 4),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2"))
)
metadata <- data.frame(class = c("A", "B"), row.names = c("s1", "s2"))
dataset <- list(data = data, metadata = metadata)
count_missing_values(dataset)

Count missing values per sample

Description

Returns the number of missing values per sample.

Usage

count_missing_values_per_sample(dataset, remove.zero = TRUE)

Arguments

Value

A named numeric vector giving the number of missing values for each sample in 'dataset$data'. If 'remove.zero' is 'TRUE', only samples with at least one missing value are returned.

Examples

data <- matrix(
  c(1, NA, 3,
    4,  5, NA),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(class = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
count_missing_values_per_sample(dataset)

Count missing values per variable

Description

Returns the number of missing values per variable.

Usage

count_missing_values_per_variable(dataset, remove.zero = TRUE)

Arguments

Value

A named numeric vector giving the number of missing values for each variable in 'dataset$data'. If 'remove.zero' is 'TRUE', only variables with at least one missing value are returned.

Examples

data <- matrix(
  c(1, 2, NA,
    4, 5, NA),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(class = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
count_missing_values_per_variable(dataset)

Create dataset

Description

Creates a dataset from a numeric matrix.

Usage

create_dataset(
  datamatrix,
  type = "undefined",
  metadata = NULL,
  description = "",
  sample.names = NULL,
  x.axis.values = NULL,
  label.x = NULL,
  label.values = NULL,
  xSet = NULL
)

Arguments

Value

A list representing a specmine dataset. The returned object contains at least the elements data, a numeric matrix with variables in rows and samples in columns; type, a character string identifying the dataset type; description, a character string describing the dataset; metadata, a data frame with one row per sample when available; labels, a list with axis and value labels when provided; and xSet, an optional object associated with LC-MS processing. This object is the standard input structure used by downstream specmine analysis functions.

Examples

datamatrix <- matrix(
  c(1.1, 2.2, 3.3, 4.4, 5.5, 6.6),
  nrow = 2,
  dimnames = list(NULL, NULL)
)
metadata <- data.frame(
  class = c("A", "B", "A"),
  row.names = c("s1", "s2", "s3")
)
create_dataset(
  datamatrix,
  type = "concentrations",
  metadata = metadata,
  sample.names = c("s1", "s2", "s3"),
  x.axis.values = c("v1", "v2"),
  label.x = "variable",
  label.values = "intensity"
)

Creates the pathway, with reactions included in the nodes

Description

Creates the pathway, with reactions included in the nodes

Usage

create_pathway_with_reactions(
  path,
  path.name,
  identified_cpds,
  nodeNames = "kegg",
  nodeTooltip = FALSE,
  map.zoom = FALSE,
  map.layout = "preset",
  map.width = NULL,
  map.height = NULL
)

Arguments

Value

A cyjShiny widget representing the pathway with reactions included in the nodes. The widget contains the pathway graph ready for interactive visualization, with identified compounds highlighted when they are present in the pathway.

Examples

## Not run: 
path <- get_MetabolitePath("hsa00010")
create_pathway_with_reactions(
  path = path,
  path.name = "hsa00010",
  identified_cpds = c("cpd:C00031", "cpd:C00022")
)

## End(Not run)

Create a dataset from peak lists

Description

Merges equivalent peaks across samples, extracts intensities, and creates a dataset object in standard package format.

Usage

dataset_from_peaks(
  sample.list,
  metadata = NULL,
  description = "",
  type = "nmr-peaks"
)

Arguments

Value

A dataset object created from the input peak lists.

Examples

sample.list <- list(
  s1 = data.frame(ppm = c(1.0, 2.0), int = c(10, 20)),
  s2 = data.frame(ppm = c(1.0, 3.0), int = c(15, 30))
)
metadata <- data.frame(class = c("A", "B"), row.names = c("s1", "s2"))
dataset <- dataset_from_peaks(sample.list, metadata = metadata)
class(dataset)

Auto-exported function: feature_selection

Description

Auto-exported function: feature_selection

Usage

feature_selection(
  dataset,
  column.class,
  method = "rfe",
  functions,
  validation = "cv",
  repeats = 5,
  number = 10,
  subsets = 2^(2:4)
)

Value

A feature selection object produced by caret. For method = "rfe", this is an rfe object; for method = "filter", this is an sbf object.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
feature_selection(dataset, "class", method = "filter", functions = caret::rfSBF)

Feature Selection Using Univariate Filters

Description

Performs feature selection using univariate filters.

Usage

filter_feature_selection(
  datamat,
  samples.class,
  functions = caret::rfSBF,
  method = "cv",
  repeats = 5
)

Arguments

Value

An sbf object.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
classes <- factor(c("A", "A", "B", "B", "A"))
filter_feature_selection(datamat, classes, caret::rfSBF)

Flat pattern filter

Description

Performs a flat pattern filter over the dataset.

Usage

flat_pattern_filter(
  dataset,
  filter.function = "iqr",
  by.percent = TRUE,
  by.threshold = FALSE,
  red.value = 0
)

Arguments

Value

A dataset object with the same overall structure as the input, where 'dataset$data' has been filtered according to the selected flat-pattern criterion. Variables failing the selection rule are removed, the remaining dataset components are preserved, and 'dataset$description' is updated to record the filtering step.

Examples

if (requireNamespace("specmine.datasets", quietly = TRUE)) {
  data(propolis, package = "specmine.datasets")
  propolis_proc = missingvalues_imputation(propolis)
  propolis_proc = flat_pattern_filter(propolis_proc, "iqr",
    by.percent = TRUE, red.value = 75)
}

Returns an object of KEGGPathway of the pathway especified in pathcode

Description

Returns an object of KEGGPathway of the pathway especified in pathcode

Usage

get_MetabolitePath(pathcode)

Arguments

Value

A KEGGPathway object corresponding to the pathway specified by pathcode. This object contains the parsed KEGG pathway structure and can be used as input to downstream graph conversion and visualization functions.

Examples

## Not run: 
path <- get_MetabolitePath("hsa00010")
class(path)

## End(Not run)

Get code, t number, full name and phylogeny of all organisms in KEGG:

Description

Get code, t number, full name and phylogeny of all organisms in KEGG:

Usage

get_OrganismsCodes()

Value

A data.frame with the KEGG T number, organism code, species names, and phylogeny for all organisms available in KEGG. Each row represents one organism and the columns are Tnumber, organismCode, speciesNames, and phylogeny.

Examples

## Not run: 
head(get_OrganismsCodes())

## End(Not run)

Get compound names from KEGG codes

Description

Returns compound names associated with KEGG compound identifiers.

Usage

get_cpd_names(kegg_codes)

Arguments

Value

A named character vector in which the values are KEGG compound identifiers and the names are the corresponding compound names. Each element of the returned vector represents one KEGG compound code matched to its primary compound name.

Examples

## Not run: 
get_cpd_names(c("C00031", "C00022"))

## End(Not run)

Auto-exported function: get_files_list_per_assay

Description

Auto-exported function: get_files_list_per_assay

Usage

get_files_list_per_assay(studyID)

Arguments

Value

A named list of data.frame objects, one per assay, each containing the sample names and corresponding file names for the specified study. Each data frame has two columns: Samples, with sample identifiers, and Files, with the associated raw or spectral data file names available in MetaboLights.

Examples

## Not run: 
assays <- get_files_list_per_assay("MTBLS1")
names(assays)

## End(Not run)

Get vector with paths numbers that occur in the given organism, named with the full path name:

Description

Get vector with paths numbers that occur in the given organism, named with the full path name:

Usage

get_metabPaths_org(org_code)

Arguments

Value

A named character vector with the pathway identifiers available for the specified organism, where each value is an organism-specific pathway code and each name is the corresponding KEGG pathway name.

Examples

## Not run: 
head(get_metabPaths_org("hsa"))

## End(Not run)

Download a complete MetaboLights study

Description

Downloads files and metadata for every assay in a public MetaboLights study.

Usage

get_metabolights_study(studyID, directory, verbose = TRUE)

Arguments

Value

Invisibly returns a list with one entry per assay containing downloaded files, metadata, and sample-file mappings.

Examples

## Not run: 
tmpdir <- tempdir()
get_metabolights_study("MTBLS1", directory = tmpdir, verbose = FALSE)

## End(Not run)

Download files for one MetaboLights assay

Description

Downloads the data files associated with one assay of a MetaboLights study.

Usage

get_metabolights_study_files_assay(studyID, assay, directory, verbose = TRUE)

Arguments

Value

Invisibly returns a character vector with downloaded file paths.

Examples

## Not run: 
tmpdir <- tempdir()
get_metabolights_study_files_assay("MTBLS1", assay = 1, directory = tmpdir, verbose = FALSE)

## End(Not run)

Get metadata for one MetaboLights assay

Description

Returns the metadata associated with the samples in one assay of a MetaboLights study and can optionally save it as a CSV file.

Usage

get_metabolights_study_metadata_assay(studyID, assay, directory = NULL)

Arguments

Value

Invisibly returns a data.frame with sample metadata for the selected assay.

Examples

## Not run: 
md <- get_metabolights_study_metadata_assay("MTBLS1", assay = 1)
head(md)

## End(Not run)

Get sample-file mapping for one MetaboLights assay

Description

Returns the sample-to-file mapping for one assay in a MetaboLights study and can optionally save it as a CSV file.

Usage

get_metabolights_study_samples_files(studyID, assay, directory = NULL)

Arguments

Value

Invisibly returns a two-column object with samples and corresponding files.

Examples

## Not run: 
get_metabolights_study_samples_files("MTBLS1", assay = 1)

## End(Not run)

Get only the paths of the organism that contain given compounds:

Description

Get only the paths of the organism that contain given compounds:

Usage

get_paths_with_cpds_org(organism_code, compounds, full.result = TRUE)

Arguments

Value

A data.frame with the pathways containing the input compounds. Each row represents one matched pathway. When full.result = TRUE, the returned data frame includes the columns pathways, ratio, compounds, and compounds_names; otherwise it contains only pathways and ratio. Row names correspond to pathway names.

Examples

## Not run: 
cpds <- c(glucose = "cpd:C00031", pyruvate = "cpd:C00022")
head(get_paths_with_cpds_org("hsa", cpds, full.result = FALSE))

## End(Not run)

Auto-exported function: get_peak_values

Description

Auto-exported function: get_peak_values

Usage

get_peak_values(samples.df, peak.val)

Arguments

Value

A numeric vector with the intensity values for the peak specified by peak.val across all samples in samples.df.

Examples

samples.df <- data.frame(
  s1 = c(10, 20),
  s2 = c(15, NA),
  row.names = c("1", "2")
)
get_peak_values(samples.df, "1")

Auto-exported function: get_samples_names_dx

Description

Extract sample names from JDX files in a folder.

Usage

get_samples_names_dx(foldername)

Arguments

Value

A character vector with the sample names extracted from the JDX files in the input folder.

Examples

dir.create("jdx_examples", showWarnings = FALSE)
file.create(file.path("jdx_examples", "sample1.dx"))
file.create(file.path("jdx_examples", "sample2.DX"))
get_samples_names_dx("jdx_examples")

Get x label

Description

Returns the x-axis label associated with the dataset.

Usage

get_x_label(dataset)

Arguments

Value

A character string giving the label of the x axis stored in 'dataset$labels$x'. If no x-axis label has been defined, the function returns an empty string.

Examples

datamatrix <- matrix(
  c(1, 2, 3, 4),
  nrow = 2,
  dimnames = list(c("10.5", "11.0"), c("s1", "s2"))
)
metadata <- data.frame(class = c("A", "B"), row.names = c("s1", "s2"))
dataset <- create_dataset(
  datamatrix,
  type = "concentrations",
  metadata = metadata,
  label.x = "ppm",
  label.values = "intensity"
)
get_x_label(dataset)

Get x values as text

Description

Returns the x values of a dataset as text.

Usage

get_x_values_as_text(dataset)

Arguments

Value

A character vector containing the variable identifiers stored in 'rownames(dataset$data)'. Each element corresponds to one row of the data matrix and represents the x-axis value or variable label associated with that feature.

Examples

datamatrix <- matrix(
  c(1, 2, 3, 4),
  nrow = 2,
  dimnames = list(c("10.5", "11.0"), c("s1", "s2"))
)
metadata <- data.frame(class = c("A", "B"), row.names = c("s1", "s2"))
dataset <- create_dataset(
  datamatrix,
  type = "concentrations",
  metadata = metadata
)
get_x_values_as_text(dataset)

Impute missing values with kNN

Description

Replace missing values using k-nearest neighbors imputation.

Usage

impute_nas_knn(dataset, k = 10, ...)

Arguments

Value

A dataset object with the same structure as the input, where missing values in 'dataset$data' have been imputed using the k-nearest neighbors method implemented in 'impute::impute.knn()'. The returned object preserves the remaining dataset components unchanged.

Examples

data <- matrix(
  c(1,  2, NA, 4,
    2,  3,  4, 5,
    3, NA,  5, 6,
    4,  5,  6, 7),
  nrow = 4,
  byrow = TRUE,
  dimnames = list(c("x1", "x2", "x3", "x4"), c("s1", "s2", "s3", "s4"))
)
dataset <- list(data = data)
impute_nas_knn(dataset, k = 2)

Impute missing values with mean

Description

Replace missing values in each variable by the mean of the observed values for that variable.

Usage

impute_nas_mean(dataset)

Arguments

Value

A dataset object with the same structure as the input, where missing values in 'dataset$data' have been replaced by the mean of the corresponding variable calculated with 'na.rm = TRUE'.

Examples

data <- matrix(
  c(1, NA, 3, 4),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2"))
)
dataset <- list(data = data)
impute_nas_mean(dataset)

Impute missing values with median

Description

Replace missing values in each variable by the median of the observed values for that variable.

Usage

impute_nas_median(dataset)

Arguments

Value

A dataset object with the same structure as the input, where missing values in 'dataset$data' have been replaced by the median of the corresponding variable calculated with 'na.rm = TRUE'.

Examples

data <- matrix(
  c(1, NA, 5, 4),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2"))
)
dataset <- list(data = data)
impute_nas_median(dataset)

Impute missing values with a constant

Description

Replace all 'NA' values in 'dataset$data' by a user-defined constant.

Usage

impute_nas_value(dataset, value)

Arguments

Value

A dataset object with the same structure as the input, where all missing values in 'dataset$data' have been replaced by 'value'. Other components of the dataset are preserved unchanged.

Examples

data <- matrix(
  c(1, NA, 3, 4),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2"))
)
dataset <- list(data = data)
impute_nas_value(dataset, 0)

Merge data and metadata

Description

Merges data and metadata into a data frame.

Usage

merge_data_metadata(
  dataset,
  samples = NULL,
  metadata.vars = NULL,
  x.values = NULL,
  by.index = FALSE
)

Arguments

Value

A 'data.frame' containing the selected data matrix, transposed so that samples are rows, combined with the selected metadata columns.

Examples

data <- matrix(
  1:6,
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = c("A", "B", "A"),
  batch = c(1, 1, 2),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
merge_data_metadata(dataset, metadata.vars = "class")

List public MetaboLights studies

Description

Returns the identifiers of public studies available in MetaboLights.

Usage

metabolights_studies_list()

Value

A character vector with public MetaboLights study identifiers.

Examples

## Not run: 
head(metabolights_studies_list())

## End(Not run)

Missing values imputation

Description

Impute missing values in a dataset using different methods.

Usage

missingvalues_imputation(dataset, method = "value", value = 5e-04, k = 5)

Arguments

Value

A dataset object with the same overall structure as the input, in which missing values in 'dataset$data' have been imputed according to the selected method. The returned object preserves the dataset components and updates the description to record the imputation step.

Examples

data <- matrix(
  c(1, NA, 3, 4, 5, NA),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
dataset <- list(data = data, description = "toy dataset")
missingvalues_imputation(dataset, method = "value", value = 0)

Multi-class summary metrics

Description

Compute overall and class-averaged performance metrics for multiclass classification models, including ROC AUC and log-loss when class probabilities are available.

Usage

multiClassSummary(data, lev = NULL, model = NULL)

Arguments

Value

A named numeric vector with overall and class-averaged classification statistics.

Examples

data <- data.frame(
  pred = factor(c("A", "B", "A", "B"), levels = c("A", "B")),
  obs = factor(c("A", "B", "B", "B"), levels = c("A", "B")),
  A = c(0.8, 0.2, 0.7, 0.3),
  B = c(0.2, 0.8, 0.3, 0.7)
)
multiClassSummary(data)

Creates the pathway wanted. If any of the given compounds is present in the pathway, it is coloured differently.

Description

Creates the pathway wanted. If any of the given compounds is present in the pathway, it is coloured differently.

Usage

pathway_analysis(
  compounds,
  pathway,
  nodeNames = "kegg",
  nodeTooltip = FALSE,
  map.zoom = FALSE,
  map.layout = "preset",
  map.width = NULL,
  map.height = NULL,
  verbose = TRUE
)

Arguments

Value

A cyjShiny widget representing the selected pathway with the input compounds highlighted when present. The returned widget can be printed or embedded in an interactive R session to inspect the pathway graph.

Examples

## Not run: 
pathway_analysis(
  compounds = c(glucose = "cpd:C00031", pyruvate = "cpd:C00022"),
  pathway = "hsa00010",
  verbose = FALSE
)

## End(Not run)

Auto-exported function: pca_analysis_dataset

Description

Auto-exported function: pca_analysis_dataset

Usage

pca_analysis_dataset(
  dataset,
  scale = TRUE,
  center = TRUE,
  write.file = FALSE,
  file.out = NULL,
  ...
)

Arguments

Value

An object of class prcomp containing PCA results. The x component stores sample scores and the rotation component stores variable loadings.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)

PCA biplot

Description

Draws a PCA biplot for a prcomp or princomp result.

Usage

pca_biplot(
  dataset,
  pca.result,
  cex = 0.8,
  legend.cex = 0.8,
  x.colors = 1,
  inset = c(0, 0),
  legend.place = "topright",
  ...
)

Arguments

Value

No return value, called for its side effect of plotting.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_biplot(dataset, pca_res)

PCA 3D biplot

Description

Creates an interactive 3D PCA biplot using rgl.

Usage

pca_biplot3D(dataset, pca.result, column.class = NULL, pcas = c(1, 2, 3))

Arguments

Value

No return value, called for its side effect of plotting.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_biplot3D(dataset, pca_res, column.class = "class")

PCA 2D k-means plot

Description

Plots PCA scores in 2D coloured by k-means cluster assignment.

Usage

pca_kmeans_plot2D(
  dataset,
  pca.result,
  num.clusters = 3,
  pcas = c(1, 2),
  kmeans.result = NULL,
  labels = FALSE,
  bw = FALSE,
  ellipses = FALSE,
  leg.pos = "right",
  xlim = NULL,
  ylim = NULL
)

Arguments

Value

A ggplot object.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_kmeans_plot2D(dataset, pca_res, num.clusters = 2)

PCA 3D k-means plot

Description

Plots PCA scores in 3D coloured by k-means cluster assignment.

Usage

pca_kmeans_plot3D(
  dataset,
  pca.result,
  num.clusters = 3,
  pcas = c(1, 2, 3),
  kmeans.result = NULL,
  labels = FALSE,
  size = 1,
  ...
)

Arguments

Value

No return value, called for its side effect of plotting.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_kmeans_plot3D(dataset, pca_res, num.clusters = 2)

PCA pairs plot with k-means clusters

Description

Creates a pairs plot of PCA scores coloured by k-means cluster assignment.

Usage

pca_pairs_kmeans_plot(
  dataset,
  pca.result,
  num.clusters = 3,
  kmeans.result = NULL,
  pcas = c(1, 2, 3, 4, 5)
)

Arguments

Value

A ggmatrix object.

Examples

datamat <- matrix(
  rnorm(30),
  nrow = 6,
  dimnames = list(paste0("x", 1:6), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_pairs_kmeans_plot(dataset, pca_res, num.clusters = 2, pcas = 1:3)

PCA pairs plot

Description

Creates a pairs plot of PCA scores.

Usage

pca_pairs_plot(
  dataset,
  pca.result,
  column.class = NULL,
  pcas = c(1, 2, 3, 4, 5),
  ...
)

Arguments

Value

A ggmatrix object.

Examples

datamat <- matrix(
  rnorm(30),
  nrow = 6,
  dimnames = list(paste0("x", 1:6), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_pairs_plot(dataset, pca_res, column.class = "class", pcas = 1:3)

Auto-exported function: pca_plot_3d

Description

Draw a 3D PCA scatter plot.

Usage

pca_plot_3d(
  dataset,
  model,
  var.class,
  pcas = 1:3,
  colors = NULL,
  legend.place = "topright",
  ...
)

Arguments

Value

A 3D scatter plot of the selected principal components, drawn for its side effects.

Examples

## Not run: 
datamat <- matrix(
  rnorm(24),
  nrow = 4,
  dimnames = list(paste0("v", 1:4), paste0("s", 1:6))
)
metadata <- data.frame(class = factor(c("A", "A", "A", "B", "B", "B")))
dataset <- list(data = datamat, metadata = metadata)
pca_model <- list(scores = prcomp(t(datamat))$x)
pca_plot_3d(dataset, pca_model, "class")

## End(Not run)

Robust PCA analysis

Description

Performs robust PCA using pcaPP::PCAgrid().

Usage

pca_robust(
  dataset,
  center = "median",
  scale = "mad",
  k = 10,
  write.file = FALSE,
  file.out = NULL,
  ...
)

Arguments

Value

An object returned by pcaPP::PCAgrid().

Examples

datamat <- matrix(
  rnorm(30),
  nrow = 6,
  dimnames = list(paste0("x", 1:6), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_robust(dataset, k = 2)

PCA 2D scores plot

Description

Creates a 2D PCA scores plot.

Usage

pca_scoresplot2D(
  dataset,
  pca.result,
  column.class = NULL,
  pcas = c(1, 2),
  labels = FALSE,
  ellipses = FALSE,
  bw = FALSE,
  pallette = 2,
  leg.pos = "right",
  xlim = NULL,
  ylim = NULL
)

Arguments

Value

A ggplot object.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_scoresplot2D(dataset, pca_res, column.class = "class")

PCA 3D scores plot

Description

Creates a static 3D PCA scores plot using scatterplot3d.

Usage

pca_scoresplot3D(dataset, pca.result, column.class = NULL, pcas = c(1, 2, 3))

Arguments

Value

No return value, called for its side effect of plotting.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_scoresplot3D(dataset, pca_res, column.class = "class")

PCA 3D scores plot using rgl

Description

Creates an interactive 3D PCA scores plot.

Usage

pca_scoresplot3D_rgl(
  dataset,
  pca.result,
  column.class = NULL,
  pcas = c(1, 2, 3),
  size = 1,
  labels = FALSE
)

Arguments

Value

No return value, called for its side effect of plotting.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_scoresplot3D_rgl(dataset, pca_res, column.class = "class")

PCA scree plot

Description

Draws a scree plot with individual and cumulative explained variance.

Usage

pca_screeplot(
  pca.result,
  num.pcs = NULL,
  cex.leg = 0.8,
  leg.pos = "right",
  lab.text = c("individual percent", "cumulative percent"),
  fill.col = c("blue", "red"),
  ylab = "Percentage",
  xlab = "Principal components",
  ...
)

Arguments

Value

No return value, called for its side effect of plotting.

Examples

datamat <- matrix(
  rnorm(20),
  nrow = 4,
  dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
)
metadata <- data.frame(class = factor(c("A", "A", "B", "B", "A")))
dataset <- list(data = datamat, metadata = metadata)
pca_res <- pca_analysis_dataset(dataset)
pca_screeplot(pca_res)

Detect peaks in 2D NMR spectra

Description

Detects local maxima in each spectrum of a 2D NMR dataset and converts the detected peaks into a 1D specmine dataset suitable for downstream analysis.

Usage

peak_detection2d(
  specmine_2d_dataset,
  baseline_thresh = NULL,
  noiseFilt = 0,
  negatives = FALSE,
  verbose = TRUE
)

Arguments

Value

A specmine dataset object as a list. The data element is a numeric matrix where rows correspond to detected 2D peak coordinates encoded as combined F1/F2 positions and columns correspond to samples. Matrix entries contain peak intensities, with missing values indicating that a given peak was not detected in a sample. The returned object also includes the original sample metadata, the dataset description, the dataset type set to "nmr-peaks", and labels describing the x-axis and data values. This output represents a reduced 1D peak table derived from the input 2D NMR spectra.

Examples

m <- matrix(c(0, 5, 0, 0), nrow = 2)
rownames(m) <- c("1.0", "1.1")
colnames(m) <- c("2.0", "2.1")
x <- list(data = list(s1 = m), metadata = data.frame(), description = "Toy")
try(peak_detection2d(x, baseline_thresh = 1, negatives = TRUE, verbose = FALSE))

Read all CSV peak files in a folder

Description

Lists CSV files in a folder and reads them into a named list of data frames.

Usage

read_csvs_folder(foldername, verbose = TRUE, ...)

Arguments

Value

A named list of data frames, one per CSV file found in the folder.

Examples

folder <- file.path(tempdir(), "peak_csvs")
dir.create(folder, showWarnings = FALSE)
utils::write.csv(data.frame(ppm = c(1, 2), int = c(10, 20)),
                 file.path(folder, "sample1.csv"), row.names = FALSE)
utils::write.csv(data.frame(ppm = c(1, 3), int = c(15, 30)),
                 file.path(folder, "sample2.csv"), row.names = FALSE)
read_csvs_folder(folder, verbose = FALSE)

Reads a data matrix from a CSV file

Description

Imports a numeric data table from a CSV file and converts it to the internal matrix structure used by the package.

Usage

read_data_csv(
  filename,
  format = "row",
  header.col = TRUE,
  header.row = TRUE,
  sep = ","
)

Arguments

Value

A numeric matrix containing the imported data in the package's internal orientation, with variables in rows and samples in columns. If format = "row", the input table is transposed before being returned. The function stops with an error if non-numeric values are found in the data table.

Examples

data_file <- tempfile(fileext = ".csv")
data_in <- data.frame(
  x1 = c(1.1, 2.2, 3.3),
  x2 = c(4.4, 5.5, 6.6),
  row.names = c("s1", "s2", "s3")
)
utils::write.csv(data_in, data_file)
read_data_csv(data_file, format = "row", header.row = TRUE)

Read JDX spectra files from a folder

Description

Reads all .dx files in a folder using readJDX::readJDX().

Usage

read_data_dx(foldername, debug = 0, verbose = TRUE)

Arguments

Value

A named list with one imported JDX object per file.

Examples

folder <- tempdir()
## Not run: 
read_data_dx(folder)

## End(Not run)

Reads a dataset from CSV files

Description

Reads a dataset from a CSV file containing the data matrix and, optionally, a second CSV file containing metadata.

Usage

read_dataset_csv(
  filename.data,
  filename.meta = NULL,
  type = "undefined",
  description = "",
  label.x = NULL,
  label.values = NULL,
  sample.names = NULL,
  format = "row",
  header.col = TRUE,
  header.row = TRUE,
  sep = ",",
  header.col.meta = TRUE,
  header.row.meta = TRUE,
  sep.meta = ","
)

Arguments

Value

An object of class dataset created from the input files. This object contains the imported data matrix and, if provided, the sample metadata, together with dataset-level information such as type, description, labels, and sample names. The returned object is intended to be used as input to downstream processing and analysis functions in the package.

Examples

data_file <- tempfile(fileext = ".csv")
meta_file <- tempfile(fileext = ".csv")

data_in <- data.frame(
  x1 = c(1.1, 2.2, 3.3),
  x2 = c(4.4, 5.5, 6.6),
  row.names = c("s1", "s2", "s3")
)
metadata_in <- data.frame(
  class = c("A", "B", "A"),
  row.names = c("s1", "s2", "s3")
)

utils::write.csv(data_in, data_file)
utils::write.csv(metadata_in, meta_file)

dataset <- read_dataset_csv(
  filename.data = data_file,
  filename.meta = meta_file,
  format = "row",
  header.row = TRUE,
  header.row.meta = TRUE
)
class(dataset)

Read a dataset from JDX files

Description

Reads all JDX spectra files in a folder, combines their intensity values into a data matrix, and optionally attaches sample metadata.

Usage

read_dataset_dx(
  folder.data,
  filename.meta = NULL,
  type = "undefined",
  description = "",
  label.x = NULL,
  label.values = NULL,
  header.col.meta = TRUE,
  header.row.meta = TRUE,
  sep.meta = ",",
  verbose = TRUE
)

Arguments

Value

A dataset object created from the spectra found in folder.data and the optional metadata file.

Examples

folder <- tempdir()
meta_file <- tempfile(fileext = ".csv")
metadata_in <- data.frame(sample = c("sample1", "sample2"), class = c("A", "B"))
utils::write.csv(metadata_in, meta_file, row.names = FALSE)
## Not run: 
read_dataset_dx(folder, filename.meta = meta_file)

## End(Not run)

Reads metadata from a CSV file

Description

Reads metadata from a CSV file and returns it as a data frame.

Usage

read_metadata(filename, header.col = TRUE, header.row = TRUE, sep = ",")

Arguments

Value

A data.frame containing the imported metadata. Rows usually correspond to samples and columns correspond to metadata variables.

Examples

meta_file <- tempfile(fileext = ".csv")
metadata_in <- data.frame(
  class = c("A", "B", "A"),
  batch = c("b1", "b1", "b2"),
  row.names = c("s1", "s2", "s3")
)
utils::write.csv(metadata_in, meta_file)
read_metadata(meta_file, header.row = TRUE)

Auto-exported function: read_ms_spectra

Description

Auto-exported function: read_ms_spectra

Usage

read_ms_spectra(
  folder.name,
  type = "undefined",
  filename.meta = NULL,
  description = "",
  prof.method = "bin",
  fwhm = 30,
  bw = 30,
  intvalue = "into",
  header.col.meta = TRUE,
  header.row.meta = TRUE,
  sep.meta = ","
)

Arguments

Value

A dataset object containing the processed LC/GC-MS spectra, the sample metadata if provided, and the associated xcms object in xSet.

Examples

## Not run: 
folder.name <- "path/to/folder/with/mzml_or_mzxml_files"
read_ms_spectra(folder.name, filename.meta = NULL)

## End(Not run)

Read multiple CSV peak files

Description

Reads multiple CSV files, one per sample, and returns them as a named list.

Usage

read_multiple_csvs(filenames, ext = ".csv", verbose = TRUE, ...)

Arguments

Value

A named list of data frames, one per input file.

Examples

f1 <- tempfile(fileext = ".csv")
f2 <- tempfile(fileext = ".csv")
utils::write.csv(data.frame(ppm = c(1, 2), int = c(10, 20)), f1, row.names = FALSE)
utils::write.csv(data.frame(ppm = c(1, 3), int = c(15, 30)), f2, row.names = FALSE)
read_multiple_csvs(c(f1, f2), ext = "", verbose = FALSE)

Import for Thermo Galactic's spc file format These functions allow to import .spc files. A detailed description of the .spc file format is available at

Description

Import for Thermo Galactic's spc file format These functions allow to import .spc files. A detailed description of the .spc file format is available at

Usage

read_spc_nosubhdr(
  filename,
  keys.hdr2data = c("fexper", "fres", "fsource"),
  keys.hdr2log = c("fdate", "fpeakpt"),
  keys.log2data = FALSE,
  keys.log2log = TRUE,
  log.txt = TRUE,
  log.bin = FALSE,
  log.disk = FALSE,
  hdr = list(),
  nosubhdr = TRUE,
  no.object = FALSE
)

Arguments

Value

A list with imported spectra information.

Author(s)

C. Beleites

References

Reference information for the SPC file format.

Examples

filename <- system.file("extdata", "toy.spc", package = "specmine")
if (nzchar(filename)) {
  read_spc_nosubhdr(filename)
}

Recursive Feature Elimination

Description

Performs recursive feature elimination on a data matrix.

Usage

recursive_feature_elimination(
  datamat,
  samples.class,
  functions = caret::rfFuncs,
  method = "cv",
  repeats = 5,
  number = 10,
  subsets = 2^(2:4)
)

Arguments

Value

An rfe object.

Examples

if (requireNamespace("randomForest", quietly = TRUE)) {
  datamat <- matrix(
    rnorm(20),
    nrow = 4,
    dimnames = list(paste0("x", 1:4), paste0("s", 1:5))
  )
  classes <- factor(c("A", "A", "B", "B", "A"))
  recursive_feature_elimination(datamat, classes, caret::rfFuncs)
}

Remove data

Description

Removes selected samples, data variables, or metadata variables from a dataset.

Usage

remove_data(
  dataset,
  data.to.remove,
  type = "sample",
  by.index = FALSE,
  rebuild.factors = TRUE
)

Arguments

Value

A dataset object with the same overall structure as the input, where the requested samples, data variables, or metadata variables have been removed according to 'type'.

Examples

data <- matrix(
  1:9,
  nrow = 3,
  dimnames = list(c("x1", "x2", "x3"), c("s1", "s2", "s3"))
)
metadata <- data.frame(class = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
remove_data(dataset, "s2", type = "sample")

Remove data variables

Description

Removes x variables from a dataset.

Usage

remove_data_variables(dataset, variables.to.remove, by.index = FALSE)

Arguments

Value

A dataset object with the same structure as the input, where the selected rows of 'dataset$data' have been removed. If no matching variables are found, the input dataset is returned unchanged and a warning is issued.

Examples

data <- matrix(
  1:9,
  nrow = 3,
  dimnames = list(c("10", "20", "30"), c("s1", "s2", "s3"))
)
metadata <- data.frame(class = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
remove_data_variables(dataset, "20")

Remove metadata variables

Description

Removes metadata variables from a dataset.

Usage

remove_metadata_variables(dataset, variables.to.remove)

Arguments

Value

A dataset object with the same structure as the input, where the selected columns have been removed from 'dataset$metadata'. If no matching metadata fields are found, the dataset is returned unchanged and a warning is issued.

Examples

data <- matrix(
  1:6,
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = c("A", "B", "A"),
  batch = c(1, 1, 2),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
remove_metadata_variables(dataset, "batch")

Remove samples

Description

Removes samples from a dataset.

Usage

remove_samples(dataset, samples.to.remove, rebuild.factors = TRUE)

Arguments

Value

A dataset object with the same structure as the input, with the selected samples removed from both 'dataset$data' and 'dataset$metadata'. If 'rebuild.factors' is 'TRUE', unused factor levels in metadata are dropped.

Examples

data <- matrix(
  1:9,
  nrow = 3,
  dimnames = list(c("x1", "x2", "x3"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = factor(c("A", "B", "A")),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
remove_samples(dataset, "s2")

Remove samples by NA metadata

Description

Removes samples with NA in a given metadata variable.

Usage

remove_samples_by_na_metadata(dataset, metadata.var)

Arguments

Value

A dataset object with the same structure as the input, where samples with missing values in the selected metadata variable have been removed.

Examples

data <- matrix(
  1:9,
  nrow = 3,
  dimnames = list(c("x1", "x2", "x3"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = c("A", NA, "B"),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
remove_samples_by_na_metadata(dataset, "class")

Remove samples by NAs

Description

Removes samples with too many missing values.

Usage

remove_samples_by_nas(dataset, max.nas = 0, by.percent = FALSE)

Arguments

Value

A dataset object with the same structure as the input, where samples whose number of missing values exceeds 'max.nas' have been removed from both 'dataset$data' and 'dataset$metadata'.

Examples

data <- matrix(
  c(1, NA, 3,
    4,  5, NA,
    7,  8, 9),
  nrow = 3,
  dimnames = list(c("x1", "x2", "x3"), c("s1", "s2", "s3"))
)
metadata <- data.frame(class = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
remove_samples_by_nas(dataset, max.nas = 0)

Remove variables by NAs

Description

Removes variables with too many missing values.

Usage

remove_variables_by_nas(dataset, max.nas = 0, by.percent = FALSE)

Arguments

Value

A dataset object with the same structure as the input, where variables whose number of missing values exceeds 'max.nas' have been removed from 'dataset$data'.

Examples

data <- matrix(
  c(1,  2, 3,
    NA, 5, NA,
    7,  8, 9),
  nrow = 3,
  dimnames = list(c("x1", "x2", "x3"), c("s1", "s2", "s3"))
)
metadata <- data.frame(class = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
remove_variables_by_nas(dataset, max.nas = 1)

Remove x values by interval

Description

Removes x values inside an interval.

Usage

remove_x_values_by_interval(dataset, min.value, max.value)

Arguments

Value

A dataset object with the same structure as the input, where all variables with x values between 'min.value' and 'max.value', inclusive, have been removed from 'dataset$data'.

Examples

data <- matrix(
  1:12,
  nrow = 4,
  dimnames = list(c("10", "20", "30", "40"), c("s1", "s2", "s3"))
)
metadata <- data.frame(class = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
remove_x_values_by_interval(dataset, 15, 35)

Spectra processing options

Description

Default options used in spectral data processing workflows.

Usage

data(spectra_options)

Format

A list containing spectral processing options.

Value

A named list of default options used by the package for spectral data processing. Each element stores a processing parameter or lookup table that controls analysis steps such as preprocessing, identification, or scoring.

Examples

data(spectra_options)
names(spectra_options)

Subset by samples and x values

Description

Selects samples and x values simultaneously.

Usage

subset_by_samples_and_xvalues(
  dataset,
  samples,
  variables = NULL,
  by.index = FALSE,
  variable.bounds = NULL,
  rebuild.factors = TRUE
)

Arguments

Value

A dataset object with the same overall structure as the input, containing only the selected samples and selected x values. The returned object includes the corresponding subset of 'dataset$data' and matching rows of 'dataset$metadata'; factor levels are rebuilt when 'rebuild.factors = TRUE'.

Examples

data <- matrix(
  1:12,
  nrow = 4,
  dimnames = list(c("10", "20", "30", "40"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = factor(c("A", "B", "A")),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
subset_by_samples_and_xvalues(dataset, samples = c(1, 3), variables = c("10", "30"))

Subset metadata

Description

Selects metadata variables to keep.

Usage

subset_metadata(dataset, variables)

Arguments

Value

A dataset object with the same structure as the input, where 'dataset$metadata' contains only the selected metadata variables.

Examples

data <- matrix(
  1:6,
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = c("A", "B", "A"),
  batch = c(1, 1, 2),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
subset_metadata(dataset, "class")

Subset random samples

Description

Selects a random subset of samples from the dataset.

Usage

subset_random_samples(dataset, nsamples)

Arguments

Value

A dataset object with the same structure as the input, containing a random subset of 'nsamples' samples.

Examples

set.seed(123)
data <- matrix(
  1:12,
  nrow = 3,
  dimnames = list(c("x1", "x2", "x3"), c("s1", "s2", "s3", "s4"))
)
metadata <- data.frame(
  class = factor(c("A", "A", "B", "B")),
  row.names = c("s1", "s2", "s3", "s4")
)
dataset <- list(data = data, metadata = metadata)
subset_random_samples(dataset, 2)

Subset samples

Description

Returns a dataset with a selected set of samples.

Usage

subset_samples(dataset, samples, rebuild.factors = TRUE)

Arguments

Value

A dataset object with the same overall structure as the input, containing only the selected samples in both 'dataset$data' and 'dataset$metadata'. If 'rebuild.factors' is 'TRUE', unused factor levels in metadata are dropped.

Examples

data <- matrix(
  c(1, 2, 3, 4, 5, 6),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = factor(c("A", "B", "A")),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
subset_samples(dataset, c("s1", "s3"))

Subset samples by metadata values

Description

Selects a set of samples by the value of a metadata variable.

Usage

subset_samples_by_metadata_values(dataset, metadata.varname, values)

Arguments

Value

A dataset object with the same structure as the input, containing only samples whose 'metadata.varname' value matches one of the requested 'values'.

Examples

data <- matrix(
  c(1, 2, 3, 4, 5, 6),
  nrow = 2,
  dimnames = list(c("x1", "x2"), c("s1", "s2", "s3"))
)
metadata <- data.frame(
  class = factor(c("A", "B", "A")),
  row.names = c("s1", "s2", "s3")
)
dataset <- list(data = data, metadata = metadata)
subset_samples_by_metadata_values(dataset, "class", "A")

Subset x values

Description

Selects rows of the dataset by x values or by row index.

Usage

subset_x_values(dataset, variables, by.index = FALSE)

Arguments

Value

A dataset object with the same structure as the input, where 'dataset$data' contains only the selected variables or row indexes.

Examples

data <- matrix(
  1:9,
  nrow = 3,
  dimnames = list(c("10", "20", "30"), c("s1", "s2", "s3"))
)
metadata <- data.frame(group = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
subset_x_values(dataset, c("10", "30"))

Subset x values by interval

Description

Selects x values within a numeric interval.

Usage

subset_x_values_by_interval(dataset, min.value, max.value)

Arguments

Value

A dataset object with the same structure as the input, where 'dataset$data' contains only variables whose x values fall between 'min.value' and 'max.value', inclusive.

Examples

data <- matrix(
  1:12,
  nrow = 4,
  dimnames = list(c("10", "20", "30", "40"), c("s1", "s2", "s3"))
)
metadata <- data.frame(group = c("A", "B", "A"), row.names = c("s1", "s2", "s3"))
dataset <- list(data = data, metadata = metadata)
subset_x_values_by_interval(dataset, 15, 35)

Summarise variable importance tables

Description

Keep the top rows of each variable-importance table produced during model comparison.

Usage

summary_var_importance(performances, number.rows)

Arguments

Value

A list of truncated variable-importance tables, one per model.

Examples

performances <- list(
  vips = list(
    rf = data.frame(Mean = c(0.9, 0.7, 0.4), row.names = c("v1", "v2", "v3")),
    svm = data.frame(Mean = c(0.8, 0.6, 0.5), row.names = c("v1", "v2", "v3"))
  )
)
summary_var_importance(performances, 2)

Train a classifier and predict new samples

Description

Train a classification model from a dataset and use the fitted model to predict the class labels of new samples.

Usage

train_and_predict(
  dataset,
  new.samples,
  column.class,
  model,
  validation,
  num.folds = 10,
  num.repeats = 10,
  tunelength = 10,
  tunegrid = NULL,
  metric = NULL,
  summary.function = caret::defaultSummary
)

Arguments

Value

A list with two elements: train.result, the fitted training object, and predictions.result, a data frame with predicted classes for new.samples.

Examples

## Not run: 
datamat <- matrix(
  rnorm(24),
  nrow = 4,
  dimnames = list(paste0("v", 1:4), paste0("s", 1:6))
)
metadata <- data.frame(class = factor(c("A", "A", "A", "B", "B", "B")))
dataset <- list(data = datamat, metadata = metadata)
new.samples <- datamat[, 1:2, drop = FALSE]
train_and_predict(dataset, new.samples, "class", model = "rpart", validation = "cv")

## End(Not run)

Train a classifier

Description

Train a classifier from a dataset object using metadata or data-derived class labels and a resampling strategy supported by caret.

Usage

train_classifier(
  dataset,
  column.class,
  model,
  validation,
  num.folds = 10,
  num.repeats = 10,
  tunelength = 10,
  tunegrid = NULL,
  metric = NULL,
  summary.function = caret::defaultSummary,
  class.in.metadata = TRUE
)

Arguments

Value

A caret training object returned by caret::train().

Examples

## Not run: 
datamat <- matrix(
  rnorm(24),
  nrow = 4,
  dimnames = list(paste0("v", 1:4), paste0("s", 1:6))
)
metadata <- data.frame(class = factor(c("A", "A", "A", "B", "B", "B")))
dataset <- list(data = datamat, metadata = metadata)
train_classifier(dataset, "class", model = "rpart", validation = "cv")

## End(Not run)

Train multiple models and compare their performance

Description

Train a set of models, collect their resampling performance, optionally compute variable importance, and store fitted models and tuning summaries.

Usage

train_models_performance(
  dataset,
  models,
  column.class,
  validation,
  num.folds = 10,
  num.repeats = 10,
  tunelength = 10,
  tunegrid = NULL,
  metric = NULL,
  summary.function = "default",
  class.in.metadata = TRUE,
  compute.varimp = TRUE
)

Arguments

Value

A list containing model performance, variable importance, full tuning results, best tuning settings, optional confusion matrices, and final fitted models.

Examples

## Not run: 
datamat <- matrix(
  rnorm(24),
  nrow = 4,
  dimnames = list(paste0("v", 1:4), paste0("s", 1:6))
)
metadata <- data.frame(class = factor(c("A", "A", "A", "B", "B", "B")))
dataset <- list(data = datamat, metadata = metadata)
train_models_performance(
  dataset,
  models = c("rpart"),
  column.class = "class",
  validation = "cv",
  compute.varimp = FALSE
)

## End(Not run)