GRASP application aims to support the geoscience community by simplifying the grain size prediction process based on Partial Least Squares (PLS) regression, utilizing data such as geochemical composition (X-Ray Fluorescence: XRF) and density scanning data (Computed Tomography: CT).
The easiest and fastest way to access the tool is directly via a web browser at: https://kathasaluib.shinyapps.io/GRASP/.
To install the latest version of GRASP, use the remotes (Csárdi et
al., 2024) package to install it from GitHub.
# Install remotes if not already installed
install.packages("remotes")
# Install GRASP from GitHub - when the repository is open
remotes::install_github("kathasal/GRASP")
# Install GRASP from GitHub - when the repository is private and the token is available
remotes::install_github("kathasal/GRASP", auth_token='token_name', upgrade='never')The package requires several other packages: if you do not have them installed, they should install automatically.
Alternativelly, you can also install the application from the ZIP file
available at Zenodo repository. The easiest way to do that is to use the
devtools package (Wickham et al., 2022).
devtools::install_local("folder_path/GRASP-main.zip", upgrade = 'never')After installation, load the package. You should see the welcome message as below.
library(GRASP)##
## -------------------------------------------------------
## Welcome to GRASP - GRAin Size Predictor!
##
## A workflow created to help predict mean grain size in sediment cores.
## Developed by Katarzyna Hasal at UiB (University of Bergen).
##
## Please, make sure your version is up to date.
## To get the latest version install from GitHub repository again.
##
## Useful Links:
## - GitHub Repo (currently private): https://github.com/kathasal/GRASP
## - online version: https://kathasaluib.shinyapps.io/GRASP/
##
## -------------------------------------------------------
##
You can then access the functions and most importantly - the application provided by the package. To open it from your R console, simply type
GRASP()and the application window should pop-up (see below). Please note that every action taken when running the GUI, will be recorded as lines of code in the console.
To use the GUI, the user must provide their data in an Excel file (.xls or .xlsx). For an example of the required format, they can refer to the example dataset (GRASP_TestDataset.xlsx), located in the data-raw folder. The file contains grain size measurements, XRF- and CT-scanning results from sediment core AST-P2-18, originally published by van der Bilt et al. (2021) and used in the current study as a natural sediment application example.
The first sheet contains the training dataset with depth, measured grain size, and corresponding geochemical and CT variables. The second sheet contains the prediction dataset comprising depth intervals for which grain size predictions are to be made, along with the associated geochemical and CT variables.
Additionally, this dataset is accessible within the package. To access
the samples that are measured for grain size (Grain size mean
(
iceland## # A tibble: 88 × 18
## Unit Material `Depth (cm)` `Grain size mean (µm)` `Grain size (phi)` K
## <chr> <chr> <dbl> <dbl> <dbl> <dbl>
## 1 Unit 1 Organic 12.5 27.2 5.20 57.4
## 2 Unit 1 Organic 22.5 26.2 5.25 65
## 3 Unit 1 Organic 32.5 26.5 5.24 54.2
## 4 Unit 1 Organic 42.5 23.9 5.39 48.9
## 5 Unit 1 Organic 52.5 30.0 5.06 55.6
## 6 Unit 1 Organic 62.5 24.4 5.36 39.8
## 7 Unit 1 Organic 72.5 30.8 5.02 64
## 8 Unit 1 Organic 82.5 23.4 5.42 42.9
## 9 Unit 1 Organic 86.5 26.5 5.24 64.1
## 10 Unit 1 Organic 88.5 13.0 6.26 226.
## # ℹ 78 more rows
## # ℹ 12 more variables: Ti <dbl>, Mn <dbl>, Fe <dbl>, Sr <dbl>, Zr <dbl>,
## # `clr-K` <dbl>, `clr-Ti` <dbl>, `clr-Mn` <dbl>, `clr-Fe` <dbl>,
## # `clr-Sr` <dbl>, `clr-Zr` <dbl>, `Density (CT greyscale)` <dbl>
And to see the high-resolution (every 0.5cm) scanning data.
iceland_predict## # A tibble: 854 × 16
## Unit Material `Depth (cm)` K Ti Mn Fe Sr Zr `clr-K`
## <chr> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
## 1 Unit 1 Organic 8.5 124. 2422. 666. 62432 329. 321. -2.14
## 2 Unit 1 Organic 9 150. 2762. 753. 68191 315. 272. -2.00
## 3 Unit 1 Organic 9.5 131. 2308. 648. 58609 299. 305 -2.05
## 4 Unit 1 Organic 10 113 1947. 542. 54787 226. 288. -2.03
## 5 Unit 1 Organic 10.5 96 1934. 564. 57000 266. 335. -2.25
## 6 Unit 1 Organic 11 79.5 1546. 422. 47223 204. 335. -2.26
## 7 Unit 1 Organic 11.5 92.2 1544. 433. 47724 155. 315. -2.06
## 8 Unit 1 Organic 12 87.5 1423 421. 43581 165. 286 -2.07
## 9 Unit 1 Organic 12.5 57.4 1062. 314. 41098 178. 302. -2.34
## 10 Unit 1 Organic 13 71.5 959. 293. 36233 174 348. -2.13
## # ℹ 844 more rows
## # ℹ 6 more variables: `clr-Ti` <dbl>, `clr-Mn` <dbl>, `clr-Fe` <dbl>,
## # `clr-Sr` <dbl>, `clr-Zr` <dbl>, `Density (CT greyscale)` <dbl>
You can see the description of the datasets in R by typing in the console ?iceland and ?iceland_predict respectively.
For the details on how to use the application please refer to the user guide and video tutorial. The package also allows to use the functions independently from the console. As we aim to focus on the application, in this section, we will present only two functionalities. The demonstration is done on the test dataset, available within the package, described in the section above.
# Calculation of grain size in logarithmic phi scale, having measurements in µm
mu_scale_data <- iceland$`Grain size mean (µm)`
phi_scale_data <- calculate_phi_scale(mu_scale_data)
df <- data.frame("File values" = iceland$`Grain size (phi)`,
"Calculated values" = phi_scale_data)
head(df)## File.values Calculated.values
## 1 5.200395 5.20
## 2 5.253574 5.25
## 3 5.236563 5.24
## 4 5.387379 5.39
## 5 5.057747 5.06
## 6 5.358932 5.36
# Calculation of CLR transform of compositional data
compositional <- names(iceland)[6:11]
compositional_data <- dplyr::select(iceland, all_of(compositional))
clr_transformed_data <- calculate_CLR(compositional_data)
head(clr_transformed_data)## calc_clr-K calc_clr-Ti calc_clr-Mn calc_clr-Fe calc_clr-Sr calc_clr-Zr
## 1 -2.330527 0.5884356 -0.6299782 4.243770 -1.2006920 -0.6710084
## 2 -2.370141 0.7283560 -0.6688395 4.095142 -1.0150993 -0.7694180
## 3 -2.441858 0.7382176 -0.8822945 3.950131 -0.9237262 -0.4404700
## 4 -2.372693 0.6237852 -0.7876984 4.020023 -1.1020487 -0.3813676
## 5 -2.452887 0.6977867 -0.6963276 4.212356 -1.1876003 -0.5733279
## 6 -2.601281 0.6893465 -0.6724720 4.028340 -0.9432658 -0.5006678
For more detailed documentation on each function, type ?function_name
in R console.
GRASP is based on a number of revised R (R Core Team, 2023) packages.
The creation of the application was aided by shiny (Chang et al.,
2024), which provides the framework for building the app. Moreover,
shinyjs (Attali, 2021) and bslib (Sievert et al., 2025) are used to
improve and customize the user interface.
We rely on the pls (Liland et al., 2024) package to build Partial
Least Squares regression models. readxl (Wickham & Bryan, 2023) and
openxlsx (Schauberger & Walker, 2024) help to handle excel files,
while ggplot2 (Wickham, Chang, et al., 2024), plotly (Sievert et
al., 2024), DT (Xie et al., 2024), gridExtra (Auguie, 2017) and
corrplot (Wei & Simko, 2024) serve for visualization purposes.
Additionally, dplyr (Wickham et al., 2023), reshape2 (Wickham, 2020)
and magrittr (Bache & Wickham, 2022) help with data transformation.
Finally, package development, testing and creation of documentation was
possible with devtools (Wickham et al., 2022), roxygen2 (Wickham,
Danenberg, et al., 2024), testthat (Wickham, 2024), knitr (Xie,
2024) and rmarkdown (Allaire et al., 2024) packages.
Allaire, J., Xie, Y., Dervieux, C., McPherson, J., Luraschi, J., Ushey, K., Atkins, A., Wickham, H., Cheng, J., Chang, W., & Iannone, R. (2024). Rmarkdown: Dynamic documents for r. https://github.com/rstudio/rmarkdown
Attali, D. (2021). Shinyjs: Easily improve the user experience of your shiny apps in seconds. https://deanattali.com/shinyjs/
Auguie, B. (2017). gridExtra: Miscellaneous functions for "grid" graphics. https://CRAN.R-project.org/package=gridExtra
Bache, S. M., & Wickham, H. (2022). Magrittr: A forward-pipe operator for r. https://magrittr.tidyverse.org
Chang, W., Cheng, J., Allaire, J., Sievert, C., Schloerke, B., Xie, Y., Allen, J., McPherson, J., Dipert, A., & Borges, B. (2024). Shiny: Web application framework for r. https://shiny.posit.co/
Csárdi, G., Hester, J., Wickham, H., Chang, W., Morgan, M., & Tenenbaum, D. (2024). Remotes: R package installation from remote repositories, including GitHub. https://remotes.r-lib.org
Liland, K. H., Mevik, B.-H., & Wehrens, R. (2024). Pls: Partial least squares and principal component regression. https://github.com/khliland/pls
R Core Team. (2023). R: A language and environment for statistical computing. R Foundation for Statistical Computing. https://www.R-project.org/
Schauberger, P., & Walker, A. (2024). Openxlsx: Read, write and edit xlsx files. https://ycphs.github.io/openxlsx/index.html
Sievert, C., Cheng, J., & Aden-Buie, G. (2025). Bslib: Custom bootstrap ’sass’ themes for shiny and rmarkdown. https://rstudio.github.io/bslib/
Sievert, C., Parmer, C., Hocking, T., Chamberlain, S., Ram, K., Corvellec, M., & Despouy, P. (2024). Plotly: Create interactive web graphics via plotly.js. https://plotly-r.com
van der Bilt, Willem G.M. et al. (2021). Late holocene canyon-carving floods in northern iceland were smaller than previously reported. Commun. Earth Environ., 2(86). https://doi.org/10.1038/s43247-021-00152-4
Wei, T., & Simko, V. (2024). Corrplot: Visualization of a correlation matrix. https://github.com/taiyun/corrplot
Wickham, H. (2020). reshape2: Flexibly reshape data: A reboot of the reshape package. https://github.com/hadley/reshape
Wickham, H. (2024). Testthat: Unit testing for r. https://testthat.r-lib.org
Wickham, H., & Bryan, J. (2023). Readxl: Read excel files. https://readxl.tidyverse.org
Wickham, H., Chang, W., Henry, L., Pedersen, T. L., Takahashi, K., Wilke, C., Woo, K., Yutani, H., Dunnington, D., & van den Brand, T. (2024). ggplot2: Create elegant data visualisations using the grammar of graphics. https://ggplot2.tidyverse.org
Wickham, H., Danenberg, P., Csárdi, G., & Eugster, M. (2024). roxygen2: In-line documentation for r. https://roxygen2.r-lib.org/
Wickham, H., François, R., Henry, L., Müller, K., & Vaughan, D. (2023). Dplyr: A grammar of data manipulation. https://dplyr.tidyverse.org
Wickham, H., Hester, J., Chang, W., & Bryan, J. (2022). Devtools: Tools to make developing r packages easier. https://devtools.r-lib.org/
Xie, Y. (2024). Knitr: A general-purpose package for dynamic report generation in r. https://yihui.org/knitr/
Xie, Y., Cheng, J., & Tan, X. (2024). DT: A wrapper of the JavaScript library DataTables. https://github.com/rstudio/DT