Package 'simplerspec'

Data frame with all measured (observed) and predicted variables.

Multiple arguments with observed (measured)-predicted pairs, specified with dplyr::vars(o = <column_name>, p = <column_name>). Column names can strings or symbols. The arguments in ... need to be named.

Character vector with package used for metrics calculation. Default is "simplerspec", which uses simplerspec::evaluate_model().

String with name for the new column that specifies the model or the outcome variable. Default is "model".

Tibble data frame containing at least the grouping column given in argument by and input spectra given in list-column column_in.

Character vector of length 1L or name/symbol that specifies the column by which groups of spectra are averaged. Default is "sample_id".

Character vector of length 1L or or name/symbol that specifies the list-column that contains the inputs spectra to be averaged. Default is "spc_rs", which are resampled spectra (i.e., resulting after preceding resample_spc() step).

Spectral data in a tibble data frame (classes "tibble_df", "tbl" and "data.frame").

Character vector of column names of list-columns to be bound into a list of data.tables

Character vector denoting column name for a unique spectrum ID. Default is "unique_id".

Character vector denoting column name for the spectrum group ID. Default is "sample_id". The group ID can later be used for plotting spectra by group (e.g. by using different colors or panels).

Data frame containing wavenumber and vip columns (numeric)

data.frame with predicted and observed data in columns.

Column that contains observed values, symbol/name or character (wrapped in "").

Column that contains predicted values, symbol/name or character (wrapped in "").

data.frame with predicted and observed data in columns.

Column that contains observed values, symbol/name or character (wrapped in "").

Column that contains predicted values, symbol/name or character (wrapped in "").

Spectral tibble (data frame) with spectral data contained in list-columns

Character vector containing names of list-columns to be extracted into a list of data.tables

Model output list returned from simplerspec::fit_pls().

Tibble that contains spectra, metadata and chemical reference as list-columns. The tibble to be supplied to spec_chem can be generated by the ⁠join_chem_spc() function⁠

Response variable as symbol or name (without quotes, no character string). The provided response symbol needs to be a column name in the spec_chem tibble.

Depreciated and replaced by response

Logical whether to perform mean centering of each spectrum column (e.g. wavenumber or wavelength) after common spectrum preprocessing. Default is center = TRUE

Logical whether to perform standard deviation scaling of each spectrum column (e.g. wavenumber or wavelength) after common spectrum preprocessing. Default is scale = TRUE

Character string stating evaluation method. Either "test_set" (default) or "resampling". "test_set" will split the data into a calibration (training) and validation (test) set, and evaluate the final model by predicting on the validation set. If "resampling", the finally selected model will be evaluated based on the cross-validation hold-out predictions.

Depreciated and replaced by evaluation_method. Default is TRUE.

Method how to to split the data into a independent test set. Default is "ken_sto", which will select samples for calibration based on Kennard-Stone sampling algorithm of preprocessed spectra. The proportion of validation to the total number of samples can be specified in the argument ratio_val. split_method = "random" will create a single random split.

Ratio of validation (test) samples to total number of samples (calibration (training) and validation (test)).

Number of component used for calculating mahalanobsis distance on PCA scores for computing Kennard-Stone algorithm. Default is ken_sto_pc = 2, which will use the first two PCA components.

Depreciated; renamed argument is ken_sto_pc.

Logical

Character specifying tuning method. Tuning method affects how caret selects a final tuning value set from a list of candidate values. Possible values are "resampling", which will use a specified resampling method such as repeated k-fold cross-validation (see argument resampling_method) and the generated performance profile based on the hold-out predictions to decide on the final tuning values that lead to optimal model performance. The value "none" will force caret to compute a final model for a predefined canditate PLS tuning parameter number of PLS components. In this case, the value supplied by ncomp_fixed' is used to set model complexity at a fixed number of components.

Character specifying resampling method. Currently, "kfold_cv" (default, performs 10-fold cross-validation), "rep_kfold_cv" (performs 5-times repeated 10-fold cross-validation), "loocv" (performs leave-one-out cross-validation), and "none" (if resampling_method = "none") are supported.

Depreciated. Use resampling_method instead.

Random seed (integer) that will be used for generating resampling indices, which will be supplied to caret::trainControl. This makes sure that modeling results are constant when re-fitting. Default is resampling_seed = 123.

Maximum number of PLS components that are evaluated by caret::train. Caret will aggregate a performance profile using resampling for an integer sequence from 1 to pls_ncomp_max

Integer of fixed number of PLS components. Will only be used when tuning_method = "none" and resampling_method = "none" are used.

Logical expression whether model evaluation graphs shall be printed

Environment where function is evaluated. Default is parent.frame.

Tibble that contains spectra, metadata and chemical reference as list-columns. The tibble to be supplied to spec_chem can be generated by the ⁠join_chem_spc() function⁠

Response variable as symbol or name (without quotes, no character string). The provided response symbol needs to be a column name in the spec_chem tibble.

Depreciated and replaced by response

Character string stating evaluation method. Either "test_set" (default) or "resampling". "test_set" will split the data into a calibration (training) and validation (test) set, and evaluate the final model by predicting on the validation set. If "resampling", the finally selected model will be evaluated based on the cross-validation hold-out predictions.

Depreciated and replaced by evaluation_method. Default is TRUE.

Method how to to split the data into a independent test set. Default is "ken_sto", which will select samples for calibration based on Kennard-Stone sampling algorithm of preprocessed spectra. The proportion of validation to the total number of samples can be specified in the argument ratio_val. split_method = "random" will create a single random split.

Ratio of validation (test) samples to total number of samples (calibration (training) and validation (test)).

Number of component used for calculating mahalanobsis distance on PCA scores for computing Kennard-Stone algorithm. Default is ken_sto_pc = 2, which will use the first two PCA components.

Depreciated; renamed argument is ken_sto_pc.

Logical

Character specifying tuning method. Tuning method affects how caret selects a final tuning value set from a list of candidate values. Possible values are "resampling", which will use a specified resampling method such as repeated k-fold cross-validation (see argument resampling_method) and the generated performance profile based on the hold-out predictions to decide on the final tuning values that lead to optimal model performance. The value "none" will force caret to compute a final model for a predefined canditate PLS tuning parameter number of PLS components. In this case, the value supplied by ncomp_fixed' is used to set model complexity at a fixed number of components.

Random seed (integer) that will be used for generating resampling indices, which will be supplied to caret::trainControl. This makes sure that modeling results are constant when re-fitting. Default is resampling_seed = 123.

Depreciated. Use resampling_method instead.

Maximum random forest trees by caret::train. Caret will aggregate a performance profile using resampling for an integer sequence from 1 to ntree_max trees.

Logical expression whether model evaluation graphs shall be printed

Environment where function is evaluated. Default is parent.frame.

Recursive list named with filename (file_id) at first level entries, where each element containing a sample measurement has nested metadata ("metadata"), spectra types (see spc_types), corresponding x-axis values (see section "Details on spectra data checks and matching"). The data list is a structural convention to organize spectra and their metadata. It follows for example the list structure returned from the Bruker OPUS binary reader simplerspec::read_opus_univ().

Character vector with the spectra types to be extracted from data list and gathered into list-columns. The spectra type names need to exactly follow the naming conventions, and the element names and contents need to be present at the second list hierarchy of data. These values are allowed:

• "spc" (default): final raw spectra after atmospheric compensation, if performed (named AB in Bruker OPUS software; results from referencing sample to reference single channel reflectance and transforming to absorbance).

• "spc_nocomp": raw spectra without atmospheric correction

• "sc_sm": Single channel reflectance spectra of the samples

• "sc_rf": Single channel reflectance spectra of the reference (background spectra)

• "ig_sm": Interferograms of the sample spectra (currently only spectra without x-axis list-columns are matched and returned)

• "ig_rf": Interferograms of the reference spectra (currently only spectra without x-axis list-columns are matched and returned)

data.frame that contains chemical values of the sample

List that contains spectral data

character of column name that defines sample_ID

Tibble that contains spectral data

Tibble that contains chemical reference values of the samples

character of column name that defines sample_ID

Tibble data frame containing spectra, x-axis values, metadata and eventual measured variables as list-columns.

Character vector of spectral list-columns to be extracted. Default is c("spc", "spc_pre") (raw and preprocessed spectra).

Character vector of length 1 denoting the column name of the measure columns. This argument is optional. Default is NULL, which does not extract an additional measure column.

Character vector of column that contains a unique spectral identifier for all spectra. Default is "unique_id".

Character vector of columns that is used assigning spectra into groups. Default is "sample_id". The group_id can be used for later plotting and thereby visually separating spectral groups into using different colors or panels.

List of spectral tibbles (data frames).

Character vector of spectral list-columns to be extracted. Default is c("spc", "spc_pre") (raw and preprocessed spectra).

Character vector of length 1 denoting the column name of the measure columns. This argument is optional. Default is NULL, which does not extract an additional measure column.

Character vector of column that contains a unique spectral identifier for all spectra. Default is "unique_id".

Character vector of columns that is used assigning spectra into groups. Default is "sample_id". The group_id can be used for later plotting and thereby visually separating spectral groups into using different colors or panels.

Model output list that is returned from simplerspec::fit_pls(). This object contains a nested list with the caret::train() object (class train), based on which VIPs at finally selected number of PLS components are computed.

Character vector of list-column name in mout$data$calibration, where spectra for bottom graph are extracted. Default is "spc_mean", which plots the mean calibration spectra after resampling.

Character string of list-column name in mout$data$calibration, where spectra for bottom graph are extracted. Default is "spc_pre", which plots the preprocessed calibration spectra after resampling.

Character string that is used to assign spectra to the same group and therefore ensures that all spectra are plotted with the same colour. Default is "sample_id"

Character string of X axis title for shared x axis of stacked graphs. Default is expression(paste("Wavenumber [", cm^-1, "]"))

Y axis title of bottom spectrum. Default is "Absorbance".

Y axis title of bottom spectrum. Default is "Preprocessed Abs.".

Double between 0 and 1 that defines transparency of spectra lines in returned graph (ggplot plot object).

Tibble that contains the first set of spectra to plot as list-column

Tibble that contains the second set of spectra (optional) to plot as list-column.

Character string describing the x axis unit. Default is "wavenumber", which will produce a graph with wavenumbers on the x axis with reversed number. If x_unit = "wavelength", the axis will be in regular order (lower wavelengths in nm on the left and higher on the right side of the axis).

Character string of list-column name in tibble where spectra of desired type are extracted to plot.

Character string of column that is used to group the spectra. Default is "unique_id". If replica spectra are present in the file and processed spectra resulting after averaging need to be plotted, it is recommend to use "sample_id" as argument to group according the sample_id column in the tibble(s) containing the spectra (spc_tbl and spc_tbl_2).

Character string used for grouping the first spectra set (spc_tbl) and producing the label text accordingly. Default is "Set 1".

Character string used for grouping the second spectra set (spc_tbl_2) and producing the label text accordingly. Default is "Set 2"

Character string for the colour of the first spectra set. Default is "black".

Character string for the colour of the first spectra set. Default is "red".

Character string or mathematical expression (use expression) for the x axis title. Default is expression(paste("Wavenumber [", cm^-1, "]")).

Character string or mathematical expression (use expression) for the y axis title. Default is "absorbance".

Double in between 0 and 1. Sets the transparency for the plotted spectra lines.

Logical whether to plot a legend for the spectra describing its name selected in arguments graph_id_1 and graph_id_2. Default is TRUE.

Tibble data frame containing spectra, x-axis values, metadata and eventual measured variables as list-columns.

List of spectral tibbles (data frames). Default is NULL (argument is not used).

Character vector of spectral list-columns to be extracted. Default is "spc" (raw spectra).

Character vector of length 1 denoting the column name of the measure columns. This argument is optional. Default is NULL, which does not extract an additional measure column.

Palette value supplied to ggplot::scale_colour_brewer(). Default is "Spectral", but you can set it to the default argument 1 (will use scale_colour_brewer(..., palette = 1)).

Sets the the order of colours in the scale that is based on a measure column. Default is -1 which reverses the scale. Argument is passed on to the function ggplot2::sclae_colour_brewer() as argument direction.

Character vector denoting column name for a unique spectrum ID. Default is "unique_id".

Character vector denoting column name for the spectrum group ID. Default is "sample_id". The group ID is used for plotting spectra by group (e.g. by using different colors or panels).

Logical that specifies whether the panel names derived from a numeric group_id column are reordered using ascending numbers. Default is TRUE.

Logical defining whether spectra are colored by the column specified by group_id.

Character (1L) defining the diverging colour scales from colorbrewer.org; see ?scale_colour_brewer for supported diverging colur types (palette argument).

Logical defining whether spectra are arranged into panels by groups specified in group_id. Default is TRUE.

Logical defining whether a legend for the group_id is plotted. Default is FALSE.

Integer vector of length 1. Defines number of columns when plotting panels (facets). Default is NULL (argument not used).

Logical defining whether panels are relabeled with custom names for spectra types. Default is TRUE. When TRUE, arguments from relabel_spc_types can be passed to plot_spc_ext (supported via the ... (ellipsis) argument)

Character vector or vector of type "expression" created by mathematical expression created by expression. Custom annotation for y-axis of spectra

Integer of length 1, from 0 to 1. Defines transparency of spectral lines. Default is 0.5 (0 is completely transparent and 1 is no transparency).

Numeric vector of length 1 specifying the width of the spectral lines. Default is 0.2.

Further arguments to be passed to plot_spc_ext. Currently, arguments of relabel_spc_types are supported.

List of model output generated from calibration step (pls_ken_stone()

Tibble of spectra after preprocessing (preprocess_spc())

Logical expression wheather only one row per sample_id returned.

Tibble that contains spectra to be preprocessed within a list-column.

Character vector of predefined preprocessing options to be applied to the spectra list-column specified in column_in. Common prefined values are stated as abbreviated preprocessing methods and options such as "sg_1_w21", where "sg" stands for Savitzky-Golay and 1 for first derivative and "w21" for a window size of 21 points.

Character vector of single list-column in spc_tbl that contain list of spectra (1 row matrix) to be processed by function supplied in select.

A character string of a custom processing function that is later parsed (produces expression in a list) and evaluated within the function preprocess_spc. The character vector argument of custom_function needs to contain "spc_raw", which is the single data table of spectra that results from binding a list of data.tables (spectra to preprocess) from the spectra list-column specified in column_in. An example for a value is "prospectr::savitzkyGolay(X = spc_raw, m = 0, p = 3, w = 9)". Optional argument. Default is NULL.

Tab delmited file from ASD software export where the first column called Wavelength contais wavelengths in nanometer and the remaining columns are sample spectra referred by an ID name provided in the first row of these columns.

Character vector containing full paths of ASD binary files to be read

Character vector. Unique identifier containing file name pasted with date and time.

Character vector containing file names and exension

Character vector containing files names without extension

List-column. List of data frames containing spectral metadata

List-column. List of wavelengths vectors (numeric).

List-column. List of data.tables containing radiance sample spectra.

List-column. List of data.tables containing reference reflectance spectra.

List-column. List of data.tables containing final reflectance spectra.

Character vector with path to file

Character vector of spectra types to extract from OPUS binary file. Default is c("spc"), which will extract the final spectra, e.g. expressed in absorbance (named AB in Bruker OPUS programs). Possible additional values for the character vector supplied to extract are "ScSm" (single channel spectrum of the sample measurement), \ code"ScRf" (single channel spectrum of the reference measurment), "IgSm" (interferogram of the sample measurment) and "IgRf" (interferogram of the reference measurement).

Logical (default TRUE) whether a message is printed when an OPUS binary file is parsed into an R list entry.

Logical whether spectra after atmospheric compensation are read with an offset of -4 bites from Bruker OPUS files. Default is FALSE.

List of character vectors containing full path names of spectra

Character vector of spectra types to extract from file. Possible values are: "spc" (AB block in Bruker Opus software), "spc_nocomp" (Spectra before final atmospheric compensation; only present if background correction has been set in Opus), "ScSm" (Single channel spectrum of the sample), "ScRf" (Single channel spectrum of the sample), "IgSm" (Interferogram of the sample), "IgRf" (Interferogram of the reference). Default is extract = c("spc").

Logical (TRUE or FALSE indicating whether files are read in parallel (multiple processors or multiple cores)). Default is parallel = FALSE. If TRUE a parallel backend needs to be registered, e.g. by using the doParallel package.

Logical whether spectra after atmospheric compensation are read with an offset of -4 bites from Bruker OPUS files. Default is FALSE.

List that contains averaged spectral information in list element MIR_mean (data.frame) and metadata in data_meta (data.frame).

logical expression (TRUE or FALSE) that specifies weather spectra shall be removed. If rm = FALSE, there will be no outlier removal

Spectra data embedded in a tibble object (classes ⁠"tbl_df", "tbl", "data.frame"⁠). The spectra tibble needs to contain at least of one of the the spectra columns spc, spc_rs, spc_mean, spc_nocomp, sc_sm, sc_rf, or spc_pre (list-columns with spectral data.tables), and wavenumbers or wavelengths (list-column with vectors of x-axis values corresponding to each spectrum). The help section "Matching spectrum type and corresponding x-axis type" describes the spectra types and corresponding x-axis types.

Character vector of length 1L or symbol/name specifying the name of list-column that contains the spectra to be resampled.

Character vector of length 1L specifying the measurement unit of the x-axis values (list-column) of the input spectra in spc_tbl. Possible values are "wavenumber" (default) or "wavelength". Wavenumber is a convenient unit of frequency in the mid-infrared spectral range, where wavelength is often used as spatial period for the visible and near-infrared range.

Numeric value of lowest wavenumber. This argument will only be used if x_unit = "wavenumber". The value serves as starting value for the new wavenumber sequence that the spectra will be resampled upon. Default value is 500 (i.e., in reciprocal centimeters).

Numeric value of highest wavenumber. This argument will only be used if ⁠x_unit = "wavenumber⁠. The value will be used as last value of the new wavenumber sequence that the spectra will be resampled upon. Default value is 4000 (i.e., in reciprocal centimeters).

Numeric value of the wavenumber increment for the new wavenumber sequence that the spectra will be resampled upon. Default value is 2 (i.e., in reciprocal centimeters).

Numeric value of lowest wavelength. This argument will only be used if x_unit = "wavelength". The value serves as starting value of the new wavenumber sequence that the spectra will be resampled upon. Default value is 350 (i.e. in nanometers).

Numeric value of highest wavelength. This argument will only be used if x_unit = "wavelength". The value will be used as last value of the new wavenumber sequence that the spectra will be resampled upon. Default value is 2500 (i.e., in nanometers).

Numeric value of the wavelength increment for the new wavenumber sequence that the spectra will be resampled upon. This argument will only be used if x_unit = "wavelength". Default value is 1 (i.e., in nanometers).

Character of "linear" (default) or "spline" with the interpolation method. "spline" uses a cubic spline to interpolate the input spectra at given x-axis values to new equispaced x-axis intervals.

Spectra as tibble objects that contain preprocessed spectra

Ratio of desired reference samples to total sample number

Number of principal components (numeric). If pc < 1, the number of principal components kept corresponds to the number of components explaining at least (pc * 100) percent of the total variance.

logical expression whether a plot (ggplot2) of sample selection for reference analysis is shown in PCA space (TRUE or FALSE).

Tibble data.frame containing spectra in list-column

List-column containing spectra, specified with column name as symbols or 1L character vector.

List-column containing x-values, specified with column name as symbols or 1L character vector.

Every n-th spectral positions to keep as 1L integer vector.

Spectral data in a tibble object (classes "tibble_df", "tbl" and "data.frame"). The spectra tibble is expected to contain at least the column spc (list-column with spectral matrices stored in a list) and wavenumbers or wavelengths (list-column that contains list of x-axis values).

Character vector that specifies column name where x-axis axis units are stored within spc_tbl. Default is "wavenumber".

Character vector that specifies which column (list-column) contains spectra to be sliced. Default is "spc".

List of numeric vectors that contains upper and lower bounds of respective regions to keep in spectra. The spectral regions outside the xvalues_cut intervals will be cut out in the output spectra.

Tibble data frame

Character vector with name of column based on which tibble is split into a list of tibbles