Skip to contents


‘ir’ is an R package that contains simple functions to import, handle and preprocess infrared spectra. Infrared spectra are stored as list columns in data frames to enable efficient storage of metadata along with the spectra and support further analyses containing other data for the same samples.

Supported file formats for import are:

  1. .csv files with individual spectra.
  2. Thermo Galactic’s .spc files with individual spectra.

Provided functions for preprocessing and general handling are:

  1. baseline correction with:
    • a polynomial baseline
    • a convex hull baseline
    • a Savitzky-Golay baseline (Lasch 2012).
  2. binning.
  3. clipping.
  4. interpolating (resampling, linearly).
  5. replacing selected parts of a spectrum by a straight line.
  6. averaging spectra within specified groups.
  7. normalizing spectra:
    • to the maximum intensity
    • to the intensity at a specific x value
    • so that all intensity values sum to 1.
  8. smoothing:
    • Savitzky-Golay smoothing
    • Fourier smoothing.
  9. computing derivatives of spectra using Savitzky-Golay smoothing.
  10. mathematical transformations (addition, subtraction, multiplication, division).
  11. computing the variance of intensity values (optionally after subtracting reference spectra).
  12. computing maxima, minima, median, and ranges of intensity values of spectra.
  13. plotting.
  14. tidyverse methods.

How to install

You can install ‘ir’ from CRAN using R via:

You can install ‘ir’ from GitHub using R via:

remotes::install_github(repo = "henningte/ir")

How to use

You can load ‘ir’ in R with:

# load ir package

# load additional packages needed for this tutorial

For brief introductions, see below and the two vignettes:

  1. Introduction to the ‘ir’ package
  2. Introduction to the irclass

Sample workflow

A simple workflow would be, for example, to baseline correct the spectra, then bin them to bins with a width of 10 wavenumber units, then normalize them so that the maximum intensity value is 1 and the minimum intensity value is 0 and then plot the baseline corrected spectra for each sample and sample type. Here’s the ‘ir’ code using the built-in sample data ir_sample_data.

ir_sample_data %>%                                      # data
  ir::ir_bc(method = "rubberband") %>%                  # baseline correction
  ir::ir_bin(width = 10) %>%                            # binning
  ir::ir_normalize(method = "zeroone") %>%              # normalization
  plot() + ggplot2::facet_wrap(~ sample_type)           # plot

Data structure

You can load the sample data with:

#> # A tibble: 58 × 7
#>    id_measurement id_sample sample_type sample_comment             klason_lignin
#>             <int> <chr>     <chr>       <chr>                      <units>      
#>  1              1 GN 11-389 needles     Abies Firma Momi fir       0.359944     
#>  2              2 GN 11-400 needles     Cupressocyparis leylandii… 0.339405     
#>  3              3 GN 11-407 needles     Juniperus chinensis Chine… 0.267552     
#>  4              4 GN 11-411 needles     Metasequoia glyptostroboi… 0.350016     
#>  5              5 GN 11-416 needles     Pinus strobus Torulosa     0.331100     
#>  6              6 GN 11-419 needles     Pseudolarix amabili Golde… 0.279360     
#>  7              7 GN 11-422 needles     Sequoia sempervirens Cali… 0.329672     
#>  8              8 GN 11-423 needles     Taxodium distichum Cascad… 0.356950     
#>  9              9 GN 11-428 needles     Thuja occidentalis Easter… 0.369360     
#> 10             10 GN 11-434 needles     Tsuga caroliniana Carolin… 0.289050     
#> # … with 48 more rows, and 2 more variables: holocellulose <units>,
#> #   spectra <named list>

ir_sample_data is an object of class ir. An Object of class ir is basically a data frame where each row represents one infrared measurement and column spectra contains the infrared spectra (one per row). This allows effectively storing repeated measurements for the same sample in the same table, as well as any metadata and accessory data (e.g. nitrogen content of the sample).

The column spectra is a list column of data frames, meaning that each cell in sample_data contains for column spectra a data frame. For example, the first element of ir_sample_data$spectra represents the first spectrum as a data frame:

# View the first ten rows of the first spectrum in ir_sample_data
ir::ir_get_spectrum(ir_sample_data, what = 1)[[1]] %>% 
#> # A tibble: 10 × 2
#>        x        y
#>    <int>    <dbl>
#>  1  4000 0.000361
#>  2  3999 0.000431
#>  3  3998 0.000501
#>  4  3997 0.000571
#>  5  3996 0.000667
#>  6  3995 0.000704
#>  7  3994 0.000612
#>  8  3993 0.000525
#>  9  3992 0.000502
#> 10  3991 0.000565

Column x represents the x values (in this case wavenumbers [cm-1]) and column y the corresponding intensity values.

How to cite

Please cite this R package as:

Henning Teickner (2022). ir: Functions to Handle and Preprocess Infrared Spectra. DOI: 10.5281/zenodo.5747170. Accessed 15 Jun 2022. Online at

Companion packages

irpeat builds on ‘ir’. irpeat provides functions to analyze infrared spectra of peat (humification indices, prediction models).


Text and figures : CC BY 4.0

Code : See the DESCRIPTION file

Data : CC BY 4.0 attribution requested in reuse. See the sources section for data sources and how to give credit to the original author(s) and the source.


We welcome contributions from everyone. Before you get started, please see our contributor guidelines. Please note that this project is released with a Contributor Code of Conduct. By participating in this project you agree to abide by its terms.


The complete data in this package is derived from Hodgkins et al. (2018) and was restructured to match the requirements of ‘ir’. The original article containing the data can be downloaded from and is distributed under the Creative Commons Attribution 4.0 International License ( The data on Klason lignin and holocellulose content was originally derived from De La Cruz, Florentino B., Osborne, and Barlaz (2016).

This packages was developed in R (R version 4.2.0 (2022-04-22 ucrt)) (R Core Team 2019) using functions from devtools (Wickham, Hester, and Chang 2019), usethis (Wickham and Bryan 2019), rrtools (Marwick 2019) and roxygen2 (Wickham et al. 2019).


De La Cruz, Florentino B., Jason Osborne, and Morton A. Barlaz. 2016. “Determination of Sources of Organic Matter in Solid Waste by Analysis of Phenolic Copper Oxide Oxidation Products of Lignin.” Journal of Environmental Engineering 142 (2): 04015076.

Hodgkins, Suzanne B., Curtis J. Richardson, René Dommain, Hongjun Wang, Paul H. Glaser, Brittany Verbeke, B. Rose Winkler, et al. 2018. “Tropical peatland carbon storage linked to global latitudinal trends in peat recalcitrance.” Nature Communications 9 (1): 3640.

Lasch, Peter. 2012. “Spectral Pre-Processing for Biomedical Vibrational Spectroscopy and Microspectroscopic Imaging.” Chemometrics and Intelligent Laboratory Systems 117 (August): 100–114.

Marwick, Ben. 2019. “rrtools: Creates a Reproducible Research Compendium.”

R Core Team. 2019. “R: A Language and Environment for Statistical Computing.” Vienna, Austria: R Foundation for Statistical Computing.

Wickham, Hadley, and Jennifer Bryan. 2019. “usethis: Automate Package and Project Setup.”

Wickham, Hadley, Peter Danenberg, Gábor Csárdi, and Manuel Eugster. 2019. “roxygen2: In-Line Documentation for R.”

Wickham, Hadley, Jim Hester, and Winston Chang. 2019. “devtools: Tools to Make Developing R Packages Easier.”