# Welch two-sample t test. Base R only; tested with R 4.6.1.
report <- function(path, x) cat(path, "=", if (is.numeric(x)) sprintf("%.17g", x) else as.character(x), "\n")

d <- read.csv("data.csv")                       # columns x (thickness) and g (line A or B)
fit <- t.test(x ~ g, data = d)                  # Welch, two-sided, 95 % interval

report("/estimate",  diff(rev(tapply(d$x, d$g, mean))))   # mean(A) - mean(B)
report("/statistic", fit$statistic)
report("/df",        fit$parameter)
report("/p_value",   fit$p.value)
