Get started with RCarb

In RCarb: Dose Rate Modelling of Carbonate-Rich Samples

¹Institute of Geography, Universität Heidelberg, Germany
²Archéosciences Bordeaux, UMR 6034, CNRS-Université Bordeaux Montaigne (France)
³Department of Geography and Geology, University of Salzburg, Salzburg (Austria)"

knitr::include_graphics("../man/figures/Logo_RCarb.svg")

Scope

Getting started with a new R package can be a very tedious business (if not to say annoying). This document was written with the intention to make your first steps as painless as possible.

Quick start with the example dataset

If you have no idea what a function does and how it works, it is always a good idea to have a closer look into the example sections of the package functions. The package 'RCarb' has one central function named model_DoseRate(). The example given in the example section in the manual will be used in the following to illustrate the central package functionality in three steps.

Load 'RCarb'

library(RCarb)

Load example data

##load example data
data("Example_Data", envir = environment())

To get a first impression on how the example dataset looks like, you call the function head() to print the first five rows of a data.frame on the terminal.

head(Example_Data)

Unfortunately, the naming of the table columns is not straightforward to understand. The good news is that each column carries additional information that can be seen in the R terminal by typing, e.g., for the column 'K' (which is the 2nd column):

attributes(Example_Data$K)

It reveals that the numbers in the column correspond to the potassium concentration and are given in '%'. Similar all other columns can be inspected.

And here the full overview

##extract attributes
attributes <- lapply(1:ncol(Example_Data), function(x) attributes(Example_Data[[x]]))
attributes <- as.data.frame(data.table::rbindlist(attributes), stringAsFactors = FALSE)

df <- cbind(COLUM = colnames(Example_Data), attributes)
knitr::kable(df)

Run dose rate modelling

Now we want to start the modelling using the data given for the first sample only.

##extract only the first row 
data <- Example_Data[1,]

##run model 
results <- model_DoseRate(
  data = data,
  DR_conv_factors = "Carb2007",
  n.MC = 10, 
  txtProgressBar = FALSE)

The function returns a terminal output along with two plots, which are mostly similar to the original graphical output provided by the 'MATLAB' program 'Carb'.

In the example above the function model_DoseRate() was called with three additional arguments, DR_conv_factors = "Carb2007", n.MC = 10, txtProgressBar = FALSE. The first argument selects the dose rate conversion factors used by 'RCarb'. The second argument limits the number of Monte Carlo runs for the error estimation to 10 and the second argument prevents the plotting of the progress bar, indicating the progression of the calculation. Both arguments were solely set to reduce calculation time and output in this vignette.

Obviously, you do not want to run each row in the input table separately to model all dose rates, so to run all the modelling for all samples in the example dataset you can call the model without subsetting the dataset first. Be careful, the calculation may take some time.

results <- model_DoseRate(
  data = Example_Data)

A note on the used dose rate conversion factors: For historical reasons 'Carb' has its own set of dose rate conversion factors, which differ slightly from values in the literature (e.g., Adamiec \& Atiken, 1998) and are used in 'RCarb' as default values. However, with 'RCarb' >= 0.1.3 you can select other dose rate conversion factors. Please type ?RCarb::Reference_Data in your R terminal for further details.

Using your own dataset

Running only the example dataset is somewhat dissatisfactory, and the usual case will be that you provide your own dataset as input. While you can enter all data directly using R, the package offers another way, using external spreadsheet software such as 'Libre Office' (or, of course, MS Excel). The procedure is sketched in the following.

Create template table

The function write_InputTemplate() was written to create a template table (a CSV-file) that can be subsequently opened and filled. Using the function ensures that your input data have the correct structure, e.g., the correct number for columns and column names.

write_InputTemplate(file = "files/RCarb_Input.csv")

The path given with the argument file can be modified as needed.

Enter own data & back import into R

Own data are added using an external spreadsheet program and then save again as CSV-file.

knitr::include_graphics(path = "files/Screenshot.png")

For re-importing, the data standard R functionality can be used.

data <- read.csv(file = "files/RCarb_Input.csv")

Model the dose rate

The final modelling does not differ from the call already show above (here without a plot output):

##run model 
results <- model_DoseRate(
  data = data,
  n.MC = 10, 
  txtProgressBar = FALSE, 
  plot = FALSE)

I don't like R

Well, then you are wrong here. However, if you are just tired of using the R terminal and you want to have a graphical user interface to interact with 'RCarb'? Surprise: We also spent countless hours to develop a shiny application called 'RCarb app', and we ship it as part of the R package 'RLumShiny'.

References {-}

Adamiec, G., Aitken, M.J., 1998. Dose-rate conversion factors: update. Ancient TL 16, 37–50. http://ancienttl.org/ATL_16-2_1998/ATL_16-2_Adamiec_p37-50.pdf

RCarb documentation built on Aug. 9, 2022, 1:05 a.m.