Code supporting the manuscript "Enhancing the Accuracy of Dynamic Time Warping by Integrating Stratigraphic Constraints into the Automated Correlation of Sedimentary Sequences"
Natural Gamma Radiation data used for this paper is stored in .csv files and is freely available through the National Offshore Petroleum Information Management System (NOPIMS, Geoscience Australia) (https://www.ga.gov.au/nopims).
The repository contains the R scripts for the DTW calculations performed and presented for the Paleoceanography and Paleoclimatology paper on the methodology of DTW custom technique.
Figure 6 can be reproduced using the U1463_U1464_DTW.R file. This file contains the DTW calculations using 5 different step patterns and 3 windowing functions, followed by the codes to create the plot. The script contains the calculation of RMSE and the codes to reproduce Supplementary Figure 1.
Figure 7 can be reproduced using the U1464_U1482_DTW.R file. This file contains the DTW calculations using 5 different step patterns and 3 windowing functions, followed by the codes to create the plot. The script contains the calculation of RMSE and the codes to reproduce Supplementary Figure 3.
To reproduce Figures 8 to 11, run the Custom_Step_Pattern.R file first. To plot Figure 8, first use the Picard1-Angel2-DTW.R file to compute DTW with the standard step pattern and then with the customized 'Stratigraphy-optimized' step pattern and 'Knowledge-based' windowing function. Individual plots can be created using the codes within the same file. Before running the density plot, run the CompareWindowBorders to create the customized window borders. To reproduce Figure 8, use Picard1-Angel2-Plot.R.
Follow the same procedure for Angel-2 and Minilya-1 sites.
To create Figure 11, run the Picard1-Angel2-Minilya1-Whitetail1-Plot.R file to create the plot.
To create Figure 4 of the standard and knowledge-based window, make sure to run the entire script of Picard1-Minilya1-DTW.R file. This performs DTW with customized step pattern and windowing function. Also, run the CompareWindowBorders to create the customized window borders. Then, run the CustomWindow.R file.
The DTW correlation based on the Hagen algorithm 'Align' can be reproduced using the Hagen_DTW.R file. Before running this file, preprocess all the NGR data for Picard-1, Angel-2, Minilya-1, and Whitetail-1 sites using their respective DTW files - Picard1-Angel2-DTW.R, Picard1-Minilya1-DTW.R and Picard1-Whitetail1-DTW.R
Shield:
This work is licensed under a Creative Commons Attribution 4.0 International License.
-
Giorgino T (2009). “Computing and Visualizing Dynamic Time Warping Alignments in R: The dtw Package.” Journal of Statistical Software, 31(7), 1-24. doi:10.18637/jss.v031.i07 https://doi.org/10.18637/jss.v031.i07.
-
Meyers, S.R. (2014). Astrochron: An R Package for Astrochronology. https://cran.r-project.org/package=astrochron
-
Signorell A (2024). DescTools: Tools for Descriptive Statistics. R package version 0.99.54, https://CRAN.R-project.org/package=DescTools.
-
Hagen, C., Creveling, J., & Huybers, P. (2024). Align: A User-Friendly App for Numerical Stratigraphic Correlation. GSA Today, 34(2). https://doi.org/10.1130/GSATG575A.1