Analysis of the reflectance spectra from paintings: classification and endmembers.

archeospec is a reporting tool for automation and reproducibility of research results. It encapsulates a particular analysis on the data gathered by spectometry sensors and automatically generates the research reports in different formats.

# install.packages("devtools")
# archeospec depends on this non CRAN external dependency
devtools::install_github("demosense/unmixR", subdir = "pkg/unmixR")
# archeospec in currently only available trough github
devtools::install_github("demosense/archeospec")

If you want to generate pdf documents you need a LaTeX compiler. If you don't have one installed, we provide two options:

• Configure the tinytex package (which is included automatically as a dependence) running the following command:

tinytex::install_tinytex()

• Install a LaTeX distribution for your OS (e.g. MiKTeX for Windows, MacTeX for macos or TeX Live in linux).

This package has been developed for managing high spectral resolution reflectance from paintings.

One objective is to locate the spectral regions with the maximum information contained in the spectral signature of all the pigments.

On the other hand, when not having previous information about the pigments, like chemical composition, the package allows to statistically discriminate between the optimal (minimum number of classes and maximum information) number of them.

It is also possible the alternative way, consisting of labelling some spectral signatures with previous information to stablish the main classes and pigments.

Finally, the package provides the contribution of the endmembers to each of the spectral signature in the samples.

Please refer to our comprehensive documentation for a complete quickstart.

We provide a direct and simple API to quickly generate report documents from data sources. It can be configured to reflect any combination of the mentioned study.

The following example generates a report, using VCA to compute 3 endmembers and kmeans to perform unsupervised classification.

# Absolute directory path containing .asd files and/or folders with more .asd files

# In windows 10
path <- "C:/Users/(username)/Documents/data/signatures"

# In linux
path <- "/home/(username)/data/signatures"

genReport(
  input_source=path,
  output="/tmp/archeospec/",
  format="html_document"
  endmembers=3,
  kmeans=T,
  seed=1000
  )

The following formats are available:

• html_document (and md)
• pdf_document (and LaTeX)
• word_document
• all (generating all previous document)

The reports may be configured to show/hide some of the plots and customize properties like colors and series names. Please refer to our documentation for a comprehensive list of parameters.

We provide an advanced API for a fine grained control of the different plots. tables and statistics. You can use it to generate your own documents via rmarkdown or to extend the results with additional analysis. Please refer to our documentation to explore the package's capabilities.

This work was carried out by the Remote Sensing Group in collaboration with the Intelligent Systems and Data Mining Research Group and the Laboratory of Archaeology, Heritage and Emerging Technologies in the University of Castilla-La Mancha. The Cueva Pintada Museum and the Cabildo of Gran Canaria have funded the work.