Error in quantile.default(x[[ncol(x)]], probs = (1 + c(-level, level))/2) : missing values and NaN's not allowed if 'na.rm' is FALSE

[Yunuuuu]

When sample is small, bootstrap samples will give NA value, which prevented the calculation of confidence interval.

library(infer)
data <- tibble::tibble(
    prop = runif(10L),
    gender = rep(c("female", "male"), each = 5L)
)
lapply(seq_len(100L), function(i) {
    boot_dist <- data %>%
        specify(prop ~ gender) %>% # alt: response = age, explanatory = season
        hypothesize(null = "independence") %>%
        generate(reps = 1000, type = "bootstrap") %>%
        calculate(stat = "diff in medians", order = c("female", "male"))
    get_ci(boot_dist)
})
#> Error in quantile.default(x[[ncol(x)]], probs = (1 + c(-level, level))/2): missing values and NaN's not allowed if 'na.rm' is FALSE

^{Created on 2023-12-22 with reprex v2.0.2}

[Yunuuuu]

I don't know directly modify following line is okay?

ci_vec <- stats::quantile(x[[ncol(x)]], probs = (1 + c(-level, level)) / 2, na.rm = TRUE) 

infer/R/get_confidence_interval.R

Line 232 in e5095f0

ci_vec <- stats::quantile(x[[ncol(x)]], probs = (1 + c(-level, level)) / 2)

Or we should convert NA value into others like 0?

[simonpcouch]

Thanks for the issue! A slightly more minimal reprex:

library(infer)

data <- data.frame(
   prop = seq(0, 1, length.out = 10),
   group = rep(c("a", "b"), each = 5L)
)

set.seed(1)
boot_dist <-
   data %>%
   specify(prop ~ group) %>%
   hypothesize(null = "independence") %>%
   generate(reps = 1000, type = "bootstrap") %>%
   calculate(stat = "diff in medians", order = c("b", "a"))

get_confidence_interval(boot_dist, .95)
#> Error in quantile.default(x[[ncol(x)]], probs = (1 + c(-level, level))/2): missing values and NaN's not allowed if 'na.rm' is FALSE

^{Created on 2023-12-22 with reprex v2.0.2}

On it. :)

[github-actions]

This issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.