Engine dedicated to statistical analysis

The StatisticEngine R6 class is a framework for managing statistical analysis on data. Data can be added as a character vector with a path to a csv file with variable names as first row and analyses names as first column or a data.frame or matrix object, where the row names are used as analyses names and the column names are used as variable names.

Note

Note that the model must be created before prediction and data must have the same number of variables as the model. Also, any pre-processing applied to the model data should be applied to the data before prediction. Note that only numeric values are accepted in data and the data column names are used as variable names and data row names are used as analyses names.

Note that the model must be created before testing and data must have the same number of variables as the model. Also, any pre-processing applied to the model data should be applied to the data before testing. Note that only numeric values are accepted in data and the data column names are used as variable names and data row names are used as analyses names.

Super class

StreamFind::Engine -> StatisticEngine

Active bindings

data: Data object.
model: Statistic model.
quantification: Quantification results.
prediction_results: Prediction results.
classification_results: Classification results.

Methods

Inherited methods

Method `new()`

Creates an R6 class StatisticEngine. Child of CoreEngine R6 class.

Usage

StatisticEngine$new(metadata = NULL, workflow = NULL, analyses = NULL)

Arguments

metadata: A named list or a Metadata object. Each list element must have length one. There is no type restrictions except for name (must be type character), author (must be type character) and date (must be class POSIXct or POSIXt). A file entry must be a character string with a full path to a file.
workflow: A Workflow object, a list of ProcessingStep objects or a character string with a full path to a rds or json file containing a Workflow.
analyses: A StatisticAnalyses class object or a character vector with full file path to .csv file with variable names as first row and analyses names as first column or a data.frame or matrix object, where the row names are used as analyses names and the column names are used as variable names.

Method `get_overview()`

Gets an overview data.frame of all the analyses.

Usage

StatisticEngine$get_overview()

Method `get_classes()`

Gets the class of each analysis.

Usage

StatisticEngine$get_classes()

Method `get_concentrations()`

Gets the concentration of each analysis.

Usage

StatisticEngine$get_concentrations()

Method `get_number_variables()`

Gets the number of variables.

Usage

StatisticEngine$get_number_variables()

Method `get_model_contributions()`

Gets the model contributions.

Usage

StatisticEngine$get_model_contributions(pcs = NULL)

Arguments

pcs: Integer vector with the principle components.

Method `add_analyses()`

Adds analyses. Note that when adding new analyses, any existing results are removed. Note that the data must have the same number and names of variables (i.e., columns) as the model.

Usage

StatisticEngine$add_analyses(analyses = NULL)

Arguments

analyses: A character vector with full file path to .csv file with variable names as first row and analyses names as first column or a data.frame or matrix object.

Returns

Invisible.

Method `add_classes()`

Adds classes to the analyses.

Usage

StatisticEngine$add_classes(classes)

Arguments

classes: A character vector with the classes.

Method `add_concentrations()`

Adds concentrations to the analyses.

Usage

StatisticEngine$add_concentrations(concentrations = NA_real_)

Arguments

concentrations: A numeric vector with the concentrations.

Method `remove_analyses()`

Removes analyses.

Usage

StatisticEngine$remove_analyses(analyses = NULL)

Arguments

analyses: A character vector with the names or numeric vector with indices of the analyses to remove.

Returns

Invisible.

Method `predict()`

Predicts the data using the model.

Usage

StatisticEngine$predict(data = NULL)

Arguments

data: Data.frame, data-table or matrix with data.

Method `test()`

Tests the model using the data.

Usage

StatisticEngine$test(data = NULL)

Arguments

data: Data.frame, data-table or matrix with data.

Method `plot_data()`

Plots the data.

Usage

StatisticEngine$plot_data(
  analyses = NULL,
  features = NULL,
  transpose = FALSE,
  interactive = TRUE,
  xLab = NULL,
  yLab = NULL,
  title = NULL,
  colorGroups = NULL,
  xTickLabelsShow = TRUE
)

Arguments

analyses: Character or numeric vector with names or indexes of analyses in the Analyses object.
features: A numeric vector with the features (columns of data matrix) to plot.
transpose: Logical, if TRUE the data is transposed (i.e., column names are used as legend).
interactive: Logical (length 1). When TRUE, the data is plotted interactively using plotly.
xLab: A string with the title for the x axis.
yLab: A string with the title for the y axis.
title: A string with the title.
colorGroups: A factor character vector with the color groups for the data.
xTickLabelsShow: Logical, if TRUE the x-axis tick labels are shown.

Method `plot_explained_variance()`

Plots the model explained cumulative variance.

Usage

StatisticEngine$plot_explained_variance(
  interactive = TRUE,
  xLab = NULL,
  yLab = NULL,
  title = NULL
)

Arguments

interactive: Logical (length 1). When TRUE, the data is plotted interactively using plotly.
xLab: A string with the title for the x axis.
yLab: A string with the title for the y axis.
title: A string with the title.

Method `plot_scores()`

Plots scores of the model.

Usage

StatisticEngine$plot_scores(
  analyses = NULL,
  interactive = TRUE,
  pcs = 1:2,
  title = NULL,
  colorGroups = NULL,
  showText = TRUE,
  showLegend = TRUE
)

Arguments

analyses: Character or numeric vector with names or indexes of analyses in the Analyses object.
interactive: Logical (length 1). When TRUE, the data is plotted interactively using plotly.
pcs: A numeric vector (length 2) with the principle components to plot.
title: A string with the title.
colorGroups: A factor character vector with the color groups for the scores.
showText: Logical (length 1), set to TRUE to show the text annotations.
showLegend: Logical (length 1). Set to TRUE to show legend.

Method `plot_residuals()`

Plots residuals of the model.

Usage

StatisticEngine$plot_residuals(
  analyses = NULL,
  interactive = TRUE,
  xLab = NULL,
  yLab = NULL,
  title = NULL
)

Arguments

analyses: Character or numeric vector with names or indexes of analyses in the Analyses object.
interactive: Logical (length 1). When TRUE, the data is plotted interactively using plotly.
xLab: A string with the title for the x axis.
yLab: A string with the title for the y axis.
title: A string with the title.

Method `plot_loadings()`

Plots model loadings.

Usage

StatisticEngine$plot_loadings(
  interactive = TRUE,
  pcs = 1:2,
  colorKey = NULL,
  title = NULL,
  showText = TRUE,
  showLegend = TRUE
)

Arguments

interactive: Logical (length 1). When TRUE, the data is plotted interactively using plotly.
pcs: A vector with the principle components to plot.
colorKey: A character vector with the color key for the loading variables.
title: A string with the title.
showText: Logical (length 1), set to TRUE to show the text annotations.
showLegend: Logical (length 1). Set to TRUE to show legend.

Method `plot_resolved_spectra()`

Plots model resolved spectra.

Usage

StatisticEngine$plot_resolved_spectra(
  interactive = TRUE,
  pcs = NULL,
  original = TRUE,
  title = NULL,
  showText = TRUE,
  showLegend = TRUE
)

Arguments

interactive: Logical (length 1). When TRUE, the data is plotted interactively using plotly.
pcs: Integer vectors with the principle component to use for categorization.
original: Logical, if TRUE the original data is plotted.
title: A string with the title.
showText: Logical (length 1), set to TRUE to show the text annotations.
showLegend: Logical (length 1). Set to TRUE to show legend.

Method `plot_contributions()`

Plots model contributions.

Usage

StatisticEngine$plot_contributions(
  interactive = TRUE,
  pcs = NULL,
  title = NULL,
  showText = TRUE,
  showLegend = TRUE
)

Arguments

interactive: Logical (length 1). When TRUE, the data is plotted interactively using plotly.
pcs: Integer vectors with the principle component to use for categorization.
title: A string with the title.
showText: Logical (length 1), set to TRUE to show the text annotations.
showLegend: Logical (length 1). Set to TRUE to show legend.

Method `plot_residual_distance()`

Plots the residual distance of the model.

Usage

StatisticEngine$plot_residual_distance(...)

Arguments

...: Additional arguments passed to the plotting function.

Method `clone()`

The objects of this class are cloneable with this method.

Usage

StatisticEngine$clone(deep = FALSE)

Arguments

deep: Whether to make a deep clone.

Note

Super class

Active bindings

Methods

Public methods

Method new()

Usage

Arguments

Method get_overview()

Usage

Method get_classes()

Usage

Method get_concentrations()

Usage

Method get_number_variables()

Usage

Method get_model_contributions()

Usage

Arguments

Method add_analyses()

Usage

Arguments

Returns

Method add_classes()

Usage

Arguments

Method add_concentrations()

Usage

Arguments

Method remove_analyses()

Usage

Arguments

Returns

Method predict()

Usage

Arguments

Method test()

Usage

Arguments

Method plot_data()

Usage

Arguments

Method plot_explained_variance()

Usage

Arguments

Method plot_scores()

Usage

Arguments

Method plot_residuals()

Usage

Arguments

Method plot_loadings()

Usage

Arguments

Method plot_resolved_spectra()

Usage

Arguments

Method plot_contributions()

Usage

Arguments

Method plot_residual_distance()

Usage

Arguments

Method clone()

Usage

Arguments

Method `new()`

Method `get_overview()`

Method `get_classes()`

Method `get_concentrations()`

Method `get_number_variables()`

Method `get_model_contributions()`

Method `add_analyses()`

Method `add_classes()`

Method `add_concentrations()`

Method `remove_analyses()`

Method `predict()`

Method `test()`

Method `plot_data()`

Method `plot_explained_variance()`

Method `plot_scores()`

Method `plot_residuals()`

Method `plot_loadings()`

Method `plot_resolved_spectra()`

Method `plot_contributions()`

Method `plot_residual_distance()`

Method `clone()`