Reputation: 13
I can extract the p-value (df2) if I define modelList in this way
responseList <- names(mtcars)[-c(4,9)]
modelList <- lapply(responseList, function(resp) {
mF <- formula(paste(resp, " ~ hp*am"))
aov(mF, data = mtcars)
})
df2 <- plyr::ldply( modelList , function(x) summary(x)[[1]][["Pr(>F)"]])
names(df2) <- c(attr(modelList[[1]]$terms, "term.labels"), "residuals")
However if I define ModelList in the following way, I can not get the result.
formula <- as.formula(paste0("cbind(", paste(names(mtcars)[-c(4,9)], collapse = ","), ") ~ hp*am"))
modelList <- aov(formula, data=mtcars)
So how can I modify df2 in order to extract the pvalues from the second defined function for modelList.
df2 <- plyr::ldply( modelList , function(x) summary(x)[[1]][["Pr(>F)"]])
names(df2) <- c(attr(modelList[[1]]$terms, "term.labels"), "residuals")
Upvotes: 1
Views: 90
Reputation: 24252
Here is one of the possible ways for extracting p-values from the second modelList
.
formula <- as.formula(paste0("cbind(", paste(names(mtcars)[-c(4,9)], collapse = ","), ") ~ hp*am"))
modelList2 <- aov(formula, data=mtcars)
df2 <- plyr::ldply( summary(modelList2), function(x) x[,"Pr(>F)"])
names(df2) <- c("", rownames(modelList2[[1]])[-1], "residuals")
df2
################
hp am hp:am residuals
1 Response mpg 1.504500e-09 4.748593e-05 0.9806460396 NA
2 Response cyl 3.974632e-11 2.997217e-04 0.0969093688 NA
3 Response disp 5.835767e-11 1.164601e-05 0.0178673463 NA
4 Response drat 9.236405e-04 1.911898e-05 0.8664040719 NA
5 Response wt 1.348501e-07 2.977105e-06 0.1913211750 NA
6 Response qsec 2.312064e-07 4.577810e-04 0.1767439339 NA
7 Response vs 2.060911e-06 9.495711e-01 0.0458841894 NA
8 Response gear 1.696824e-01 1.405025e-09 0.0002293635 NA
9 Response carb 3.077405e-07 3.622291e-02 0.1902486095 NA
Upvotes: 1