Skip to contents

Simulate synthetic pXRF compositional data (Dirichlet)

Source: R/taphonomy.R
morie_taphonomy_simulate_pxrf.Rd

Synthetic data for testing/calibration ONLY – never a substitute for real comparanda, and never write the output to inst/extdata. Portable X-ray fluorescence yields elemental concentrations that are strictly non-negative and closed (a composition on the simplex), so Gaussian noise is the wrong model; this samples from a Dirichlet distribution instead. A control profile (natural soil/bone matrix) and a treatment profile (quicklime – heavily skewed to calcium) let you exercise the causal / compositional pipeline before real scans exist.

Usage

morie_taphonomy_simulate_pxrf(
  n,
  condition = c("control", "treatment"),
  elements = c("Ca", "P", "Fe", "Sr", "Pb", "Zn"),
  alpha = NULL,
  seed = NULL,
  as_ppm = FALSE,
  total_ppm = 1e+06
)

Arguments

n

Number of samples (rows).

condition

"control" (natural) or "treatment" (lime).

elements

Character vector of element names (default Ca, P, Fe, Sr, Pb, Zn).

alpha

Optional Dirichlet concentration vector (one per element). Defaults encode the control vs lime profiles; required if you pass custom elements.

seed

Optional RNG seed.

as_ppm

If TRUE, scale proportions to parts-per-million summing to total_ppm; else return proportions summing to 1.

total_ppm

Total for the ppm scaling (default 1e6).

Value

A data.frame: one column per element, plus condition and lime_treatment (1 = treatment). Attributes elements, alpha, and synthetic = TRUE are attached.

Details

Dirichlet variates are drawn via normalised Gamma (\(X_i \sim \mathrm{Gamma}(\alpha_i, 1)\), \(Y = X / \sum X\)) so no extra package is needed.

See also

morie_taphonomy_clr, morie_taphonomy_ilr

Examples

head(morie_taphonomy_simulate_pxrf(5, "treatment", seed = 1))
#>          Ca          P          Fe           Sr           Pb           Zn
#> 1 0.9160276 0.06498899 0.009709022 8.530101e-03 3.378336e-04 4.064105e-04
#> 2 0.9245194 0.05901579 0.010998388 3.742419e-03 1.585031e-03 1.389422e-04
#> 3 0.9266993 0.05568073 0.013866444 7.319622e-06 3.433476e-04 3.402894e-03
#> 4 0.9489989 0.04162588 0.006514353 7.521032e-05 6.500858e-08 2.785601e-03
#> 5 0.9189031 0.03840343 0.034643436 7.957533e-03 9.251798e-05 1.313421e-08
#>   condition lime_treatment
#> 1 treatment              1
#> 2 treatment              1
#> 3 treatment              1
#> 4 treatment              1
#> 5 treatment              1