pcdiags.plt: Diagnostic Plots for Principal Curve Analysis

In pcurve: Principal curve analysis

Description

A menu of 12 plots for diagnosis of results from principal curve analysis, pcurve

Usage

pcdiags.plt(zz, xx, pch = 1, graphics = TRUE)

Arguments

zz	an object of class principal curve, being the value of the funtion pcurve.
xx	data.frame or matrix of explanatory (environmental) variables to be used in constrained pcs.
pch	symbol to be used in plots.
graphics	a logical argument of menu indicating whether a graphics menu should be used. Currently unused.

Details

Produces a menu of 12 (or thirteen if xx is not missing) options. Once a selection is made, return to the menu by left-mouse clicking on the plot.

0. Exit

1. Residuals plots for each variable on the PC (by the internal function pcres1.plt)

2. Absolute residuals plot for each variable on the PC (by the internal function pcres2.plt)

3. QQ normal residuals plot for each variable (by the internal function pcqqnorm.plt)

4. QQ chi-squared quantile residuals plot (by the internal function pcchisq.plt)

5. Response plot and residual plot for each variable (by the internal function pcresid.plt)

6. Differenced locations: Plot of distances between consecutive locations on the PC (by the internal function pcfinder.plt)

7. Response plots for each variable along the PC (by the internal function pcresp.plt)

8. Flip plots: Plot of the PC projected onto a bi-plot of the first two principal coordinates, showing fitted locations of the variables on the PC. Left-mouse click to scroll through biplots of other principal coordinate combinations. (Right-mouse-click to return to the menu) (Using the internal function pcflip.plt)

9. Fix curve: a utility to break the curve in up to two places (by left mouse-clicks), re-order the segments and rerun the PC analysis with a new start. (using the internal function finder)

10. Scatterplots of Eclidean and Bray-Curtis distances against the PC. (using the internal function pcdists.plt)

11. Histograms of Eclidean and Bray-Curtis distances against the PC. (using the internal function pchist.plt)

12. A toggle to use Case numbers or symbols in plots

13. Env. vars. vs Gradient: if xx is not missing, plots of distance along the PC and explanatory variables (using the internal function pcenv.plt)

Value

Produces plots

Author(s)

R port by Chris Walsh [e-mail] from S+ library by Glenn De'ath [e-mail].

References

De'ath, G. 1999 Principal Curves: a new technique for indirect and direct gradient analysis. Ecology 80, 2237–2253.

pcurve documentation built on May 2, 2019, 4:20 p.m.