CODEWITHSUNDEEP
r
Tomas
Tomas

Reputation: 59435

Why does plot(Vectorize(function)) work but curve(Vectorize(function)) does not?

Why does plot(Vectorize(f)), where f is a function, work, but curve(Vectorize(f)) does not? This inconsistency bothers me a bit. Here is a full reproducable example:

set.seed(123)
N <- 9
X <- rnorm(N)
a <- 0.2
b <- 0.8
sigma <- 0.7
eps <- rnorm(N, 0, sigma)
Y <- a + b * X + eps


cor_p_value <- function(new_Y1)
{
Y[1] <- new_Y1
cor.test(X,Y)$p.value
}

lm_F_p_value <- function(new_Y1)
{
Y[1] <- new_Y1
anova(lm(Y ~ X))$Pr[1]
}

plot(Vectorize(cor_p_value), xlim = c(-50,50), lwd = 3)
curve(Vectorize(lm_F_p_value), col = "red", lwd = 3, lty = 2, add = TRUE)
# Error in curve(Vectorize(lm_F_p_value), col = "red", lwd = 3, lty = 2,  : 
# 'expr' must be a function, or a call or an expression containing 'x'

So for curve, I must do this annoying workaround:

cu <- Vectorize(lm_F_p_value)
curve(cu, col = "red", lwd = 3, lty = 2, add = TRUE)

But why? Vectorize returns the same function, plot takes it, curve does not. All these functions are from the base built-in R packages, so I expect them to cooperate and be consistent. Probably a bug?

Upvotes: 2

Views: 67

Answers (1)

akrun
akrun

Reputation: 886938

According to ?curve

expr - The name of a function, or a call or an expression written as a function of x which will evaluate to an object of the same length as x.

curve(Vectorize(lm_F_p_value)(x), col = "red", lwd = 3, lty = 2, 
       add = TRUE, ylab = "y")

-output

enter image description here

Upvotes: 3

Related Questions