jwiley / jwileymisc Goto Github PK

I note that JWileymisc imports the lsmeans package. However, lsmeans is about to be deprecated. The newer package emmeans replaces it and works in much the same way. There are vignettes and utilities to convert lsmeans scripts and workspaces.

I suggest that you update to emmeans as soon as possible, This is made perhaps more urgent by the fact that I received a message that lsmeans is failing some tests due to increased stringency in the tests regarding the use of Suggests packages. Until I saw that a new package imports lsmeans, I was prepared to answer that this would be a good time to formally deprecate lsmeans, rather than fix the package to pass the tests.

Hi,
This came up during revdep checks for lavaan. In the test-distributions.R file (in tests/testthat), you now have (lines 65-67)

x <- as.data.frame(scale(mtcars[, 1:3]))
x[1, 1] <- NA
expect_warning(m <- testDistribution(x, distr = "mvnormal", use = "fiml"))

However, the last expression runs without producing a warning, and therefore, this 'test' fails, as it was expecting a warning. Could you please adapt this line, and update JWileymisc on CRAN? (It would otherwise prevent lavaan from being updated).
Thanks,
Yves.

Hi,

I'm trying to reproduce your code from your book:

set.seed(1234)
d2 <- data.table(x = rnorm(500))
d2[, y := ifelse(abs(x) > 1, x, -x)]
plot_grid(
testdistr(d2$x, plot = FALSE)$Density,
testdistr(d2$y, plot = FALSE, varlab = "Y")$Density,
ncol = 2)

By which I got an error:

could not find function "testdistr"

I have installed and attached the "JWileymisc" package and all the other packages.
I'm wondering whether that's because I couldn't install those updates:

Rcpp 1.0.4.6 1.0.5
processx 3.4.2 3.4.3
mnormt 2.0.0 2.0.1
nloptr 1.2.2.1 1.2.2.2

My R version is 4.0.2. Do you have any suggestions for that?

x <- data.table(a = 1:5)
x$y <- scale(6:10)
winsorizor(x$y)
Error: is.vector(d) || is.matrix(d) || is.data.frame(d) is not TRUE

Correlation of perfect 1.00 is shown as .00 in SEMSummary correlation matrix plot. Colouration is correct though.