This function previously did not accommodate tibbles for the
data argument of wqs_full_perm, but now
data can be a data.frame object or a
tibble.
Added a utils.R script with a function
get_legend2() to replace the malfunctioning
cowplot::get_legend() call in the function
wqspt_plot. In cowplot v1.1.3,
get_legend() returns an error if the legend position is
anywhere besides on the right of the plot. Since wqspt_plot
extracts a legend positioned on the bottom of a plot to be included in
the plot output if InclKey = TRUE, this was previously
throwing an error.
Added a gwqs_hpc function to utils.R
along with its necessary hidden functions imported from the
gWQS package to add an option to specify a number of
workers for parallel processes for wqs_pt and
wqs_full_perm. The gwqs function in the
gWQS package (v3.0.5) uses as many parallel processes as
there are cores detected, which can be problematic with high-performance
computing (HPC) environments. HPC schedulers allocate a specific number
of cores, and if gwqs tries to use more than that number,
it will terminate the HPC job. Therefore, gwqs_hpc was
created to allow for the number of parallel processes to be specified,
thereby avoiding this problem.
Added ... arguments to wqs_pt and
wqs_full_perm that ensure that the additional arguments
passed to gwqs_hpc are passed to every iteration of that
function, whereas previously those additional arguments only passed to
the main WQS regression in wqs_full_perm.
Added the arguments LegendWidthIn and
LegendHeightIn to wqspt_plot to control the
bottom legend width and height, respectively.
Added some examples to the documentation for
wqspt_plot to better illustrate its use.
Replaced all b1_constr arguments with
b_constr to match the change in the name of this argument
in the latest version of the gWQS package
(v3.0.5).
Replaced URL links to referenced papers with doi.org links.
Added additional examples of papers using the WQSPT method.