Write a simulation function (submit an R script) Write your own function, called

Question

Write a simulation function (submit an R script) Write your own function, called bs(). It will produce a 95% bootstrap confidence interval for a mean. The user will provide a vector, and the number of repetitions for the simulation (nrep). You're writing your own function, so do not reference other bootstrap functions that exist in other packages.(e.g., there is a boot function in a bootstrap package - don't use this!) Your function should: Draw a sample the same size as the vector, selecting FROM the vector itself, with replacement. Compute the mean of the vector, and store it in another vector, Repeat the steps above n times. When done, compute the interval as the 2.5% and 97.5% quantiles of the vector of means. Round to 3 digits to the right of the decimal point, Print the interval on the console so that it looks something like this: 95% bootstrap interval for the mean: -0.309 to 0.073 Return the vector of means to the user.

Explanation / Answer

# x is a single quantitative sample
# B is the desired number of bootstrap samples to take

boot.mean = function(x,N) {
        n = length(x)
        boot.samples = matrix( sample(x,size=n*N,replace=TRUE), N, n)
        boot.statistics = apply(boot.samples,1,mean)
        interval <- quantile(boot.statistics, c(0.025, 0.975))
        print(paste( "95% bootstrap interval for the mean: ", interval[1], " to ", interval[2]))
        return(boot.statistics)
}

Related Questions

Navigate

• Browse (All)
• Browse W
• Subjects

---

---

Integrity-first tutoring: explanations and feedback only — we do not complete graded work. Learn more.