plot.lmcal, plot.nlscal {quantchem} | R Documentation |
Plot various plots commonly used to illustrate callibration.
plot.lmcal(x, type = c("curve", "residuals", "chronologic", "qqplot", "cook", "optimization"), trend = TRUE, confidence = TRUE, prediction = TRUE, lines = c(1, 2, 3, 4), colors = c(1, 1, 1, 1, 1), xlab = NULL, ylab = NULL, ...) plot.nlscal(x, type = c("curve", "residuals", "chronologic", "qqplot"), trend = TRUE, lines = c(1, 2, 3, 4), colors = c(1, 1, 1, 1, 1), xlab = NULL, ylab = NULL, ...)
x |
An object of 'lmcal' or 'nlscal' class, respectively |
type |
Type of plots: curve - calibration plots, residuals
- residuals vs. x values (common residual plot), chronologic
- residuals in chronologic order (for autocorrelation detection),
qqplot - quantile-quantile plots of residuals,
cook - cook's distances, optimization - plots of
weighting and Box-Cox optimization
|
trend |
Logical: should be the trend of residuals marked (obtained by 'loess') |
confidence |
Logical: if the confidence interval should be plotted |
prediction |
Logical: if the prediction interval should be plotted |
lines |
Line types of the plot |
colors |
Colors of the plot |
xlab |
Label of x |
ylab |
Label of y |
... |
Additional arguments, currently ignored |
Lukasz Komsta
x = rep(1:10,5) y = jitter(sqrt(x)) fit = lmcal(x,y) par(mfrow=c(2,2)) plot(fit,type="curve") plot(fit,type="residuals") plot(fit,type="optimization") par(mfrow=c(1,1)) # Low level plotting, useful for customizing plot! plot(x,y) lines(fit$graph$grid,fit$graph$fitted$p1,lty=2) lines(fit$graph$grid,fit$graph$fitted$p2,lty=2)