Create the \(\Theta\) matrix with the shared uncertainties that can used as input in, e.g., AgeS_Computation and Age_OSLC14 which is used for the covariance matrix \(\Sigma\) (Combès & Philippe, 2017)
create_ThetaMatrix(input, output_file = NULL, sigma_s, ...)character or data.frame (optional): input data frame or file connection
to import a CSV-file with the needed information. If nothing is provided the function returns
an input template. The argument output_file can be used to write this input template to the file
system
character (optional): file path for the output CSV-file, the field separator
is hard set to ",". Please use utils::write.table for more flexibility.
numeric (required): named character with values for systematic uncertainties. Those values
are lab-specific. Can be set to NULL to remove systematic uncertainties. The order of the named
vector is not important, but the naming!
Note: some of the uncertainties have a unit, please check details.
further arguments that can be passed to utils::read.table (for the CSV-file import)
A symmetric \(Theta\) matrix or if input is missing, a data.frame with an input
template
The function intends to ease the creation of the \(Theta\) matrix, which cannot be
created straight forward, e.g., base R functions such as stats::cov.
The relationship between the covariance matrix \(Sigma\) and \(Theta\) is given with
$$\Sigma_ij = A_i * A_j * \Theta_ij$$
For details see Combès & Philippe, 2017 and Guérin et al. (2021).
Input modes
The function supports two different operation modes:
input is left empty: the function returns a data.frame template that can be used as input (the option output_file works as well)
input is fed with a data.frame or a character (file path), the \(\Theta\) matrix is returned
Input format
The function expects either a CSV-file or a data.frame as input. To create template you can
run the function leaving the argument input empty (see example). Please note the format
of the input table (data.frame) needs to kept as specified in the template.
The following table lists the meaning of the columns:
| COLUMN | DESCRIPTION | UNIT |
SAMPLE_ID | sample name | - |
DR_BETA_K | standard error beta-dose rate K | Gy/ka |
DR_BETA_U | standard error beta-dose rate U | Gy/ka |
DR_BETA_Th | standard error beta-dose rate Th | Gy/ka |
DR_GAMMA_K | standard error gamma-dose rate K | Gy/ka |
DR_GAMMA_U | standard error gamma-dose rate U | Gy/ka |
DR_GAMMA_Th | standard error gamma-dose rate Th | Gy/ka |
DR_GAMMA_TOTAL | standard error total gamma-dose rate | Gy/ka |
DR_TOTAL | total dose rate | Gy/ka |
DR_TOTAL_X | standard error total dose rate | Gy/ka |
Note: All columns can be set to 0 or NA but no column must be left empty! If a value > 0 is provided
for DR_GAMMA_TOTAL this value is taken and values in, e.g., DR_GAMMA_K are discarded (set to 0)!
Systematic uncertainties
The following table provides information on the named argument
that can be provided via the argument sigma_s. Missing values are not allowed, all
values must be set.
| ARGUMENT | DESCRIPTION | UNIT |
s_betaK | relative uncertainty K concentration | - |
s_betaU | relative uncertainty U concentration | - |
s_betaTh | relative uncertainty Th concentration | - |
s_gammaK | relative uncertainty K concentration | - |
s_gammaU | relative uncertainty U concentration | - |
s_gammaTh | relative uncertainty Th concentration | - |
s_gammaDR | relative uncertainty gamma-dose rate | - |
s_CAL | relative uncertainty beta-source calibration | - |
s_intDR | absolute uncertainty internal dose rate | Gy/ka |
0.1.0
Kreutzer, S., Guérin, G., 2024. create_ThetaMatrix(): Create Theta Matrix. Function version 0.1.0. In: Christophe, C., Philippe, A., Kreutzer, S., Guérin, G., Baumgarten, F.H., Frerebeau, N., 2024. BayLum: Chronological Bayesian Models Integrating Optically Stimulated. R package version 0.3.2. https://CRAN.r-project.org/package=BayLum
Combès, B., Philippe, A., 2017. Bayesian analysis of individual and systematic multiplicative errors for estimating ages with stratigraphic constraints in optically stimulated luminescence dating. Quaternary Geochronology 39, 24–34. doi:10.1016/j.quageo.2017.02.003
Guérin, G., Lahaye, C., Heydari, M., Autzen, M., Buylaert, J.-P., Guibert, P., Jain, M., Kreutzer, S., Lebrun, B., Murray, A.S., Thomsen, K.J., Urbanova, P., Philippe, A., 2021. Towards an improvement of optically stimulated luminescence (OSL) age uncertainties: modelling OSL ages with systematic errors, stratigraphic constraints and radiocarbon ages using the R package BayLum. Geochronology 3, 229—245. doi:10.5194/gchron-3-229-2021
##(1) return template data.frame (no file output)
create_ThetaMatrix()
#> [create_ThetaMatrix()] 'input' missing, input template returned!
#> SAMPLE_ID DR_BETA_K DR_BETA_U DR_BETA_TH DR_GAMMA_K DR_GAMMA_U DR_GAMMA_TH
#> 1 NA NA NA NA NA NA NA
#> DR_GAMMA_TOTAL DR_TOTAL DR_TOTAL_X
#> 1 NA NA NA
if (FALSE) {
##(2) return template as data.frame + file
file_path <- tempfile(fileext = ".csv")
create_ThetaMatrix(output_file = file_path )
##NOT RUNNING EXAMPLE for sigma_s
calc_ThetaMatrix(...,
sigma_s = c(
s_betaK = 0.010,
s_betaU = 0.007,
s_betaTh = 0.006,
s_gammaK = 0.010,
s_gammaU = 0.007,
s_gammaTh = 0.006,
s_gammaDR = 0.05,
s_CAL = 0.020,
s_intDR = 0.030))
}