Function for plotting the cross-validated covariates traces of a sbh object.
Plot the cross-validated modal trace curves of covariate importance and covariate usage of the
pre-selected covariates specified by user at each iteration of the peeling sequence
(inner loop of our PRSP or PRGSP algorithm).
plot_trace(object,
main = NULL,
xlab = "Box Mass",
ylab = "Covariate Range (centered)",
toplot = object$cvfit$cv.used,
center = TRUE,
scale = FALSE,
col.cov,
lty.cov,
lwd.cov,
col = 1,
lty = 1,
lwd = 0.5,
cex = 0.5,
add.caption = FALSE,
text.caption = NULL,
device = NULL,
file = "Covariate Trace Plots",
path = getwd(),
horizontal = FALSE,
width = 8.5,
height = 8.5, ...)Object of class sbh as generated by the main function sbh.
Character vector. Main Title.
Defaults to NULL.
Character vector. X-axis label.
Defaults to "Box Mass".
Character vector. Y-axis label.
Defaults to "Covariate Range (centered)".
Numeric vector. Which of the pre-selected covariates to plot (in reference to the original index of covariates).
Defaults to covariates used for peeling.
Logical scalar. Shall the data be centered?.
Defaults to TRUE.
Logical scalar. Shall the data be scaled?
Defaults to FALSE.
Integer vector. Line color for the covariate importance curve of each selected covariate.
Defaults to vector of colors of length the number of selected covariates.
The vector is reused cyclically if it is shorter than the number of selected covariates.
Integer vector. Line type for the covariate importance curve of each selected covariate.
Defaults to vector of 1's of length the number of selected covariates.
The vector is reused cyclically if it is shorter than the number of selected covariates.
Integer vector. Line width for the covariate importance curve of each selected covariate.
Defaults to vector of 1's of length the number of selected covariates.
The vector is reused cyclically if it is shorter than the number of selected covariates.
Integer scalar. Line color for the covariate trace curve.
Defaults to 1.
Integer scalar. Line type for the covariate trace curve.
Defaults to 1.
Numeric scalar. Line width for the covariate trace curve.
Defaults to 0.5.
Numeric scalar. Symbol expansion used for titles, captions, and axis labels.
Defaults to 0.5.
Logical scalar. Should the caption be plotted?.
Defaults to FALSE.
Character vector of caption content.
Defaults to NULL.
Graphic display device in {NULL, "PS", "PDF"}.
Defaults to NULL (standard output screen).
Currently implemented graphic display devices are "PS" (Postscript) or "PDF" (Portable Document Format).
File name for output graphic. Defaults to "Covariate Trace Plots".
Absolute path (without final (back)slash separator). Defaults to working directory path.
Logical scalar. Orientation of the printed image.
Defaults to FALSE, that is potrait orientation.
Numeric scalar. Width of the graphics region in inches.
Defaults to 8.5.
Numeric scalar. Height of the graphics region in inches.
Defaults to 8.5.
Generic arguments passed to other plotting functions.
Invisible. None. Displays the plot(s) on the specified device.
This work made use of the High Performance Computing Resource in the Core Facility for Advanced Research Computing at Case Western Reserve University. This project was partially funded by the National Institutes of Health NIH - National Cancer Institute (R01-CA160593) to J-E. Dazard and J.S. Rao.
The trace plots limit the display of traces to those only covariates that are used for peeling. If centered, an horizontal black dotted line about 0 is added to the plot.
Due to the variability induced by cross-validation and replication, it is possible that more than one covariate be used for peeling at a given step. So, for simplicity of the trace plots, only the modal or majority vote trace value (over the folds and replications of the cross-validation) is plotted.
The top plot shows the overlay of covariate importance curves for each covariate. The bottom plot shows the overlay of covariate usage curves for each covariate. It is a dicretized view of covariate importance.
Both point to the magnitude and order with which covariates are used along the peeling sequence.
Dazard J-E. and Rao J.S. (2018). "Variable Selection Strategies for High-Dimensional Survival Bump Hunting using Recursive Peeling Methods." (in prep).
Rao J.S., Huilin Y. and Dazard J-E. (2018). "Disparity Subtyping: Bringing Precision Medicine Closer to Disparity Science." (in prep).
Diaz-Pachon D.A., Saenz J.P., Dazard J-E. and Rao J.S. (2018). "Mode Hunting through Active Information." (in press).
Diaz-Pachon D.A., Dazard J-E. and Rao J.S. (2017). "Unsupervised Bump Hunting Using Principal Components." In: Ahmed SE, editor. Big and Complex Data Analysis: Methodologies and Applications. Contributions to Statistics, vol. Edited Refereed Volume. Springer International Publishing, Cham Switzerland, p. 325-345.
Yi C. and Huang J. (2017). "Semismooth Newton Coordinate Descent Algorithm for Elastic-Net Penalized Huber Loss Regression and Quantile Regression." J. Comp Graph. Statistics, 26(3):547-557.
Dazard J-E., Choe M., LeBlanc M. and Rao J.S. (2016). "Cross-validation and Peeling Strategies for Survival Bump Hunting using Recursive Peeling Methods." Statistical Analysis and Data Mining, 9(1):12-42.
Dazard J-E., Choe M., LeBlanc M. and Rao J.S. (2015). "R package PRIMsrc: Bump Hunting by Patient Rule Induction Method for Survival, Regression and Classification." In JSM Proceedings, Statistical Programmers and Analysts Section. Seattle, WA, USA. American Statistical Association IMS - JSM, p. 650-664.
Dazard J-E., Choe M., LeBlanc M. and Rao J.S. (2014). "Cross-Validation of Survival Bump Hunting by Recursive Peeling Methods." In JSM Proceedings, Survival Methods for Risk Estimation/Prediction Section. Boston, MA, USA. American Statistical Association IMS - JSM, p. 3366-3380.
Dazard J-E. and J.S. Rao (2010). "Local Sparse Bump Hunting." J. Comp Graph. Statistics, 19(4):900-92.