calibPlane {transmem}

Calculates regression plane for external standard calibration.

Description

A bivariated regression plane for external standard calibration is calculated to later convert signals into concentration values. It differs from calibCurve in the number of explanatory variables, 2 in this case. This function is useful when some interference effect is being considered such as the magnification of the interest species signal due to the presence of another (known) species in the same sample.

Usage

calibPlane(plane, badpoint = NULL, plot = TRUE, lines = 13, theta = -30,
  phi = 40, xlab = "Species 1", ylab = "Species 2", zlab = "Signal",
  pch = 18, cex = 2)

Arguments

plane

Data frame of numeric vectors named 'Conc', 'Conc.S' and 'Signal'. The vectors must contain the concentrations of the main species (the one whose concentration in the samples is to be known) and the secondary species (the interferent), and the standard's signals, respectively.

badpoint

Numeric vector with the points to be ignored in the regresion. This allows the easy elimination of outliers without losing the stored measurement information.

plot

Logical. If TRUE, the default, the calibration data is plotted.

lines

Number of lines to use in the mesh of the plane in the plot.

theta

Azimuthal angle at which the plane is visualized.

phi

Altitude angle at which the plane is visualized.

xlab

Label for X axis (main species concentration).

ylab

Label for Y axis (secondary species concentration).

zlab

Label for Z axis (response).

pch

Plotting symbols available in R.

cex

The size of pch symbols.

Details

A linear method (i.e lm()) is applied to obtain the regression equation. The user must verify model assumptions such as normal distribution of residuals.

Value

Model of the calibration plane

Author(s)

Cristhian Paredes, [e-mail]

Eduardo Rodriguez de San Miguel, [e-mail]

Examples

  data(planelithium)
  planeModel <- calibPlane(plane = planelithium)
  summary(planeModel$model)