Scribd
Upload
4 views2 pages

Assignment 4

Uploaded by

fapic87459
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
4 views2 pages

Assignment 4

Uploaded by

fapic87459
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Assignment - 4

Data Scienc with R

Deadline: 14 July 2023, 11:59 pm


Instructions: Read from previous Assignments
Problems:
1. Construct a matrix A which is 1000x1000. Find the norm of every column of the matrix using a loop
and then also using function sapply. Which is faster?

2. Improve the efficiency of the following code that calculates 𝐴𝐵𝑡 𝑥 for matrices and �, � and vector �.

n <- 1000
m <- 1000
A <- matrix(runif(n*m), nrow = n, ncol = m)
B <- matrix(rnorm(n*m), nrow = n, ncol = m)
x <- runif(m)
ABtx <- (A %*% t(B)) %*% x

3. Load file ques3.Rdata. This object contains a 200 × 200 matrix mat. Write R code that calculates

[det(𝐴)]1/𝑝 ⋅ 𝑝! ⋅ (2.7)𝑝
ans =
𝑝𝑝 ⋅ trace(𝐴)
where for a matrix 𝐴, the determinant is the product of its eigenvalues and the trace is the sum of its
eigenvalues. The final answer should be stored in ans and the last line of the code should be

ans <- ....

HINT: you can use function eigen(mat, only.values = TRUE) to calculate the eigenvalues of mat.

4. Improve the efficiency of the following code.

n <- 50
m <- 1e3
A <- matrix(runif(n*m), nrow = n, ncol = m)
# p_vec will store p_i eventually
p_vec <- numeric(length = m)
# running a loop for each column

1
# to find the norm (the numerator)
for(k in 1:m)
{
p_vec[k] <- norm(A[ ,k], type = "2")
}
# divide by the sum
p_vec <- p_vec/sum(p_vec)
# choosing column
chosen <- sample(1:m, size = 1, prob = p_vec)

5. Consider the following function autoreg below. (It is your job to figure out what the function does.)
Write and submit another R function autoreg_fast that is much faster than the above. (You CANNOT
use Rcpp here.)

# n = integer
# rho = number in (-1,1)
autoreg <- function(n, rho)
{
out <- 0
for(t in 2:n)
{
error <- rnorm(1)
error <- rho*out[t-1] + error
out <- c(out, error)
}
return(out)
}

6. Write an R function called sel_sums that does the following tasks in the following order:

a. Takes an argument mat which is a square matrix. You may assume that the user will always input a
square matrix. Let the size of mat be 𝑝 × 𝑝.

b. Randomly chooses a number, 𝑠, between 1 and 𝑝 (inclusive), with equal probability.

c. Calls a C++ function, sumsC, that returns the sum of the first 𝑠 columns of mat.

d. Returns the output obtained from the sumsC function.

You might also like