Type:	Package
Title:	Spectroscopy Analysis Using the Tidy Data Philosophy
Version:	0.1.0
Maintainer:	Marcel Ferreira <marcel.ferreira@unesp.br>
Description:	Enables the analysis of spectroscopy data such as infrared ('IR'), Raman, and nuclear magnetic resonance ('NMR') using the tidy data framework from the 'tidyverse'. The 'tidyspec' package provides functions for data transformation, normalization, baseline correction, smoothing, derivatives, and both interactive and static visualization. It promotes structured, reproducible workflows for spectral data exploration and preprocessing. Implemented methods include Savitzky and Golay (1964) "Smoothing and Differentiation of Data by Simplified Least Squares Procedures" <doi:10.1021/ac60214a047>, Sternberg (1983) "Biomedical Image Processing" https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=1654163, Zimmermann and Kohler (1996) "Baseline correction using the rolling ball algorithm" <doi:10.1016/0168-583X(95)00908-6>, Beattie and Esmonde-White (2021) "Exploration of Principal Component Analysis: Deriving Principal Component Analysis Visually Using Spectra" <doi:10.1177/0003702820987847>, Wickham et al. (2019) "Welcome to the tidyverse" <doi:10.21105/joss.01686>, and Kuhn, Wickham and Hvitfeldt (2024) "recipes: Preprocessing and Feature Engineering Steps for Modeling" https://CRAN.R-project.org/package=recipes.
License:	MIT + file LICENSE
Encoding:	UTF-8
LazyData:	true
URL:	https://github.com/marceelrf/tidyspec
BugReports:	https://github.com/marceelrf/tidyspec/issues
RoxygenNote:	7.2.3
Depends:	R (≥ 4.1.0)
Imports:	crayon, dplyr, ggplot2, glue, plotly, purrr, recipes, readr, readxl, rlang, scales, signal, tibble, tidyr, tidyselect, timetk
Suggests:	gridExtra, knitr, rmarkdown, tidyverse
VignetteBuilder:	knitr
NeedsCompilation:	no
Packaged:	2025-06-10 12:57:32 UTC; madra
Author:	Marcel Ferreira [aut, cre], Willian Zambuzzi [dtc], Julia Moraes [dtc, ctb], Emerson Araujo Alves dos Santos [ctb]
Repository:	CRAN
Date/Publication:	2025-06-12 14:20:02 UTC

CoHAspec Dataset

Description

A dataset containing spectral absorbance measurements for different concentrations of CoHA.

Usage

CoHAspec

Format

A data frame with 6 rows and 5 columns:

Wavenumber

Numeric. The spectral wavenumber (cm-1).

CoHA01

Numeric. Absorbance values for CoHA at 1 mM Cobalt concentration.

CoHA025

Numeric. Absorbance values for CoHA at 2.5 mM Cobalt concentration.

CoHA05

Numeric. Absorbance values for CoHA at 5 mM Cobalt concentration.

CoHA100

Numeric. Absorbance values for CoHA at 10 mM Cobalt concentration.

Source

de Almeida GS, Ferreira MR, da Costa Fernandes CJ, et al. Development of cobalt (Co)-doped monetites for bone regeneration. J Biomed Mater Res. 2024; 112(1):e35319. <doi:10.1002/jbm.b.35319>

Examples

data(CoHAspec)
head(CoHAspec)

Check the Currently Set Wavenumber Column

Description

This function checks which column has been set as the default wavenumber column. If no column has been set, it prints a message prompting the user to define one.

Usage

check_wn_col(show_val = FALSE)

Arguments

Value

If 'show_val = TRUE', returns the name of the wavenumber column (character). Otherwise, it only prints a message.

Examples

set_spec_wn("Wavenumber")  # Set the wavenumber column
check_wn_col()             # Check which column is set
check_wn_col(show_val = TRUE)  # Check and return the column name

Usage Example and Test

Description

Function to demonstrate the use of implemented functions

Usage

demo_rolling_ball(verbose = TRUE)

Arguments

Value

A list containing two elements:

simple

Results from the simple rolling ball method

morphology

Results from the mathematical morphology method

Examples

# Run the demonstration
demo_results <- demo_rolling_ball()

# Access results
simple_method <- demo_results$simple
morphology_method <- demo_results$morphology

Plot Rolling Ball Results

Description

Convenience function to visualize the results

Usage

plot_rolling_ball(
  result,
  title = "Rolling Ball Baseline Correction",
  x_values = NULL
)

Arguments

Value

No return value, called for side effects (creates a plot)

Rolling Ball Baseline Correction

Description

Implements the rolling ball baseline correction method

Usage

rolling_ball(x, wm, ws = 0)

Arguments

Value

A list with three components:

baseline

Numeric vector containing the estimated baseline values

corrected

Numeric vector containing the baseline-corrected signal (original - baseline)

original

Numeric vector containing the original input signal

Examples

# Example with simulated data
x <- seq(1, 100, by = 1)
y <- sin(x/10) + 0.1*x + rnorm(100, 0, 0.1)
result <- rolling_ball(y, wm = 10)
plot(x, y, type = "l", col = "blue", main = "Rolling Ball Correction")
lines(x, result$baseline, col = "red", lwd = 2)
lines(x, result$corrected, col = "green")
legend("topright", c("Original", "Baseline", "Corrected"),
       col = c("blue", "red", "green"), lty = 1)

Enhanced Rolling Ball with Mathematical Morphology

Description

More sophisticated version of rolling ball using mathematical morphology concepts

Usage

rolling_ball_morphology(x, radius, smooth = TRUE)

Arguments

Value

A list with three components:

baseline

Numeric vector containing the estimated baseline values using morphological operations

corrected

Numeric vector containing the baseline-corrected signal (original - baseline)

original

Numeric vector containing the original input signal

Set the Default Wavenumber Column

Description

Defines a default wavenumber column name to be used by other functions in the package when 'wn_col' is not explicitly provided.

Usage

set_spec_wn(col_name)

Arguments

Value

Invisibly returns the assigned column name.

See Also

[spec_select()]

Examples

set_spec_wn("Wavenumber")

Baseline Smoothing

Description

Auxiliary function to smooth the baseline using moving average

Usage

smooth_baseline(baseline, ws)

Arguments

Value

Smoothed numeric vector of the same length as input

Convert Absorbance Data to Transmittance

Description

This function converts absorbance data to transmittance using the formula T = 10^{(2 - A)}, where A is the absorbance and T is the transmittance.

Usage

spec_abs2trans(.data, wn_col = NULL)

Arguments

Value

A 'tibble' with the converted transmittance data, containing the wavelength column and the numeric transmittance columns. Any rows with infinite values are removed.

Extract Rolling Ball Baseline from Spectral Data

Description

This function extracts the rolling ball baseline from spectral data within a specified wavelength range. It returns only the baseline, not the corrected data.

Usage

spec_bl_rollingBall(
  .data,
  wn_col = NULL,
  wn_min = NULL,
  wn_max = NULL,
  wm,
  ws = 0,
  is_abs = TRUE
)

Arguments

Value

A 'tibble' with the baseline data, containing the wavelength column and the baseline for each numeric column.

References

Baseline estimation performed using a custom rolling ball implementation.

Apply Rolling Ball Baseline Correction to Spectral Data

Description

This function applies a rolling ball baseline correction to spectral data within a specified wavelength range. It allows for correction of either absorbance or transmittance data.

Usage

spec_blc_rollingBall(
  .data,
  wn_col = NULL,
  wn_min = NULL,
  wn_max = NULL,
  wm,
  ws = 0,
  is_abs = TRUE
)

Arguments

Value

A 'tibble' with the baseline-corrected spectral data, containing the wavelength column and the corrected numeric columns.

References

Baseline estimation performed using a custom rolling ball implementation.

Apply Differentiation to Spectral Data

Description

This function applies numerical differentiation to spectral data, allowing for the calculation of the first or higher-order differences.

Usage

spec_diff(.data, wn_col = NULL, degree = 1)

Arguments

Value

A 'tibble' with the differentiated spectral data, containing the wavelength column and the differentiated numeric columns. If 'degree' is 0, the original data is returned.

Filter spectral data by wavenumber range

Description

This function filters the spectral dataset based on a specified wavenumber ('wn') range. It requires the wavenumber column to be previously set using [set_spec_wn()]. If 'wn_min' and/or 'wn_max' are provided, the data will be filtered accordingly. If neither is provided, the original dataset is returned unchanged.

Usage

spec_filter(.data, wn_min = NULL, wn_max = NULL)

Arguments

Value

A filtered data frame based on the wavenumber column.

Examples

set_spec_wn("Wavenumber")
spec_filter(CoHAspec, wn_min = 500, wn_max = 1800)

Normalize Spectral Data to the [0, 1] Range

Description

This function normalizes the numeric spectral data in each column to the [0, 1] range, preserving the wavelength column.

Usage

spec_norm_01(.data, wn_col = NULL)

Arguments

Value

A 'tibble' with the normalized spectral data, containing the wavelength column and the normalized numeric columns.

Normalize Spectral Data to a Specified Range

Description

This function normalizes the numeric spectral data in each column to a specified range [min, max], preserving the wavelength column.

Usage

spec_norm_minmax(.data, wn_col = NULL, min = 0, max = 1)

Arguments

Value

A 'tibble' with the normalized spectral data, containing the wavelength column and the normalized numeric columns.

Standardize Spectral Data to Unit Variance

Description

This function standardizes the numeric spectral data in each column to have a mean of 0 and a standard deviation of 1 (unit variance), while preserving the wavelength column.

Usage

spec_norm_var(.data, wn_col = NULL)

Arguments

Value

A 'tibble' with the standardized spectral data, containing the wavelength column and the standardized numeric columns.

Perform Principal Component Analysis (PCA) on Spectral Data

Description

This function computes a Principal Component Analysis (PCA) on spectral data, excluding the wavenumber column from the analysis.

Usage

spec_pca(.data, wn_col = NULL, scale = TRUE, center = TRUE)

Arguments

Value

A 'prcomp' object containing the PCA results, including principal components, standard deviations, and loadings.

Examples

set_spec_wn("Wavenumber")
pca_result <- spec_pca(CoHAspec)
summary(pca_result)

Scree plot for PCA results

Description

Creates a customizable scree plot based on a 'prcomp' object showing variance explained by each component.

Usage

spec_pca_screeplot(
  pca,
  n = 10,
  show_labels = TRUE,
  show_cumulative = TRUE,
  bar_color = "steelblue",
  line_color = "darkred",
  show_kaiser = FALSE,
  title = "Scree Plot",
  subtitle = NULL,
  accuracy = 1
)

Arguments

Value

A ggplot2 scree plot object.

Examples

pca <- prcomp(USArrests, scale. = TRUE)
spec_pca_screeplot(pca, n = 4)
spec_pca_screeplot(pca, show_kaiser = TRUE, bar_color = "darkblue")

Compute Wavenumber Contributions to Principal Components

Description

This function calculates the contribution of each wavenumber to the principal components (PCs) in a PCA result. Contributions are computed as the squared loadings multiplied by 100.

This function calculates the contribution of each wavenumber to the principal components (PCs) in a PCA result. Contributions are computed as the squared loadings multiplied by 100.

Usage

spec_pca_wn_contrib(PCA)

spec_pca_wn_contrib(PCA)

Arguments

Details

The function extracts the PCA loadings (rotation matrix) and computes the squared values of each loading, scaled to percentage values. This helps interpret the importance of each wavenumber in defining the principal components.

The function extracts the PCA loadings (rotation matrix) and computes the squared values of each loading, scaled to percentage values. This helps interpret the importance of each wavenumber in defining the principal components.

Value

A tibble containing the wavenumber column and the percentage contribution of each wavenumber to each principal component.

A tibble containing the wavenumber column and the percentage contribution of each wavenumber to each principal component.

Examples

pca_result <- spec_pca(CoHAspec)
wn_contrib <- spec_pca_wn_contrib(pca_result)
print(wn_contrib)



pca_result <- spec_pca(CoHAspec)
wn_contrib <- spec_pca_wn_contrib(pca_result)
print(wn_contrib)

Read Spectral Data from Various File Formats

Description

This function reads spectral data from a file, automatically detecting the format and using either 'readr' or 'readxl' functions. It also sets the specified wavenumber column as the default using 'set_spec_wn()'.

Usage

spec_read(file, wn_col, ...)

Arguments

Details

The function automatically determines the file format based on the extension: - '.txt', '.csv', '.tsv', and '.csv2' are read using 'readr' functions. - '.xls' and '.xlsx' are read using 'readxl::read_excel()'.

If the specified 'wn_col' does not exist in the data, an error is returned.

Value

A tibble containing the spectral data.

Examples

  file_path <- system.file("extdata", "CoHAspec.csv", package = "tidyspec")
  df <- spec_read(file_path, wn_col = "Wavenumber")
  check_wn_col() # Verifica se a coluna de wavenumber está definida

Select Specific Columns in a Spectral Data Frame

Description

This function selects user-specified columns from a spectral dataset, always ensuring that the wavenumber column ('wn_col') is included, unless explicitly excluded.

Usage

spec_select(.data, ...)

Arguments

Value

A data frame containing the selected columns.

See Also

[dplyr::select()], [set_spec_wn()]

Create a Custom Plot for Spectral Data

Description

This function generates a customizable plot for spectral data, allowing for the selection of plot type (absorbance or transmittance), x-axis direction, and plot geometry (points or lines).

Usage

spec_smartplot(
  .data,
  wn_col = NULL,
  xdir = c("reverse", "standard"),
  geom = c("point", "line"),
  xmin = NULL,
  xmax = NULL,
  alpha = 0.8,
  type = c("absorbance", "transmittance")
)

Arguments

Value

A 'ggplot' object representing the customized spectral plot (absorbance or transmittance as a function of wavelength/wavenumber).

Create an Interactive Plot for Spectral Data using Plotly

Description

This function generates an interactive Plotly plot for spectral data, allowing for the selection of plot type (absorbance or transmittance), x-axis direction, and plot geometry (points or lines).

Usage

spec_smartplotly(
  .data,
  wn_col = NULL,
  type = c("absorbance", "transmittance"),
  xdir = c("reverse", "standard"),
  geom = c("point", "line"),
  xmin = NULL,
  xmax = NULL,
  alpha = 0.8
)

Arguments

Value

A 'plotly' object representing the interactive spectral plot.

Apply Smoothing to Spectral Data Using a Moving Average

Description

This function applies a moving average smoothing to numeric spectral data using a specified window size and polynomial degree, while preserving the wavelength column.

Usage

spec_smooth_avg(.data, wn_col = NULL, window = 15, degree = 2)

Arguments

Value

A 'tibble' with the smoothed spectral data, containing the wavelength column and the smoothed numeric columns.

Apply Savitzky-Golay Smoothing to Spectral Data

Description

This function applies Savitzky-Golay smoothing to numeric spectral data using a specified window size, polynomial order, and differentiation degree, while preserving the wavelength column.

Usage

spec_smooth_sga(.data, wn_col = NULL, window = 15, forder = 4, degree = 0)

Arguments

Value

A 'tibble' with the smoothed spectral data, containing the wavelength column and the smoothed numeric columns.

Convert Spectral Data from Transmittance to Absorbance

Description

This function converts transmittance data to absorbance using the formula 'A = 2 - log10(T)', where 'T' is the transmittance. It also filters out any infinite values resulting from the transformation, while preserving the wavelength column.

Usage

spec_trans2abs(.data, wn_col = NULL)

Arguments

Value

A 'tibble' with the converted absorbance data, containing the wavelength column and the absorbance numeric columns.