Faux Cell Data Set for Testing & Demonstration — generate_faux

This is a synthetic data set intended for testing and demonstration. Function generate_faux_cell() simulates the faux cell data (note: in the future, it is planned to parameterize thefuncion) and object faux_cell is an instance of this dataset generated first time it is used.

The data set resembles the chondro data set but is entirely synthetic.

generate_faux_cell()

faux_cell

Format

The object faux_cell is a hyperSpec object that contains 875 Raman-like spectra, allocated to three groups/regions in column region: the matrix/background, the cell and the cell nucleus. Each spectrum is composed of 300 data points. The spectrum of each region is unique and simple, with a single peak at a particular frequency and line width. Poisson noise has been added. The data is indexed along the x and y dimensions, simulating data collected on a grid.

An object of class hyperSpec of dimension 875 x 4 x 300.

Author

Claudia Beleites, Bryan A. Hanson

Examples

set.seed(1)
faux_cell <- generate_faux_cell()

faux_cell
#> hyperSpec object
#>    875 spectra
#>    4 data columns
#>    300 data points / spectrum
#> wavelength: Delta * tilde(nu)/cm^-1 [numeric] 602 606 ... 1798 
#> data:  (875 rows x 4 columns)
#>    1. x: x position [numeric] -11.55 -10.55 ... 22.45 
#>    2. y: y position [numeric] -4.77 -4.77 ... 19.23 
#>    3. region:  [factor] matrix matrix ... matrix 
#>    4. spc: intensity (arbitrary units) [matrix, array300] 54 75 ... 108 

plot(sample(faux_cell, 10), stacked = TRUE)

# Plot mean spectra
fc_groups <- aggregate(faux_cell, faux_cell$region, mean_pm_sd)
plotspc(fc_groups,
  stacked = ".aggregate",
  col = c("red", "green", "blue"), fill = ".aggregate"
)

mapcols <- c(cell = "aquamarine", matrix = "aliceblue", nucleus = "dodgerblue")
plotmap(faux_cell, region ~ x * y, col.regions = mapcols)

# PCA
pca <- prcomp(faux_cell)
plot(pca)

loadings <- decomposition(faux_cell, t(pca$rotation), scores = FALSE)
plot(loadings[1:5], stacked = TRUE)

plot(pca$x[, 2], pca$x[, 3],
  xlab = "PC 1", ylab = "PC 2",
  bg = mapcols[faux_cell$region], col = "black", pch = 21
)