30+ R Interview Questions And Answers

The four main data structures are vectors, matrices, lists, and data frames. Vectors are a sequence of elements of the same type, matrices are 2-dimensional arrays of the same type, lists can contain elements of different types, and data frames are tables where each column has values of one type.

You can use the spread() function from the 'tidyr' package by specifying the key-value pairs to spread the data frame to a wider format.

You can use graphical methods like boxplots or scatter plots, or calculate statistical metrics like the Interquartile Range (IQR) and then check for values that fall outside of 1.5*IQR below the first quartile or above the third quartile.

S3 and S4 are two object-oriented system classes in R. S3 is a simpler, more flexible system used for implementing basic object-oriented features, while S4 is a more formal method that allows for more rigorous class definitions.

The base R plotting system is simple and easy to use, but limited in customization. The lattice system is good for creating multi-panel plots. 'ggplot2' is part of the tidyverse and allows for complex, multi-layered graphics built using a consistent grammar.

First, load 'ggplot2' with library(ggplot2), then you can create a scatter plot using the ggplot() function followed by the geom_point() function.

library(ggplot2)
ggplot(data = df, aes(x = var1, y = var2)) + geom_point()

Sure, here's a custom function to calculate the mean of a numeric vector, but exclude any NA values:

calc_mean <- function(x) {
  mean(x, na.rm = TRUE)
}

A closure in R is a function along with the environment in which it was created. It allows for keeping state between function calls.

make_adder <- function(n) {
  function(x) n + x
}

adder_10 <- make_adder(10)
adder_10(5)  # This will return 15

Lazy evaluation is a feature in R where arguments are not evaluated until they are actually used. It can save time and memory when dealing with large datasets.

f <- function(a) {
  print('This will be printed!')
  return(a * 2)
}
f(sys.sleep(5))  # The argument is not evaluated until needed, so no delay occurs

To create reproducible random number generation, use the set.seed() function with a specific integer before generating random numbers. This ensures the random numbers can be replicated.

Vectorization is the ability to perform operations on entire vectors without the need for explicit loops. This is not only concise but also more efficient computationally.

Rmarkdown is a framework for writing reproducible reports with R code embedded. 'knitr' is an engine used to convert Rmarkdown files into HTML, PDF, or other formats. You write the document in Rmarkdown syntax, embed R code chunks, and use 'knitr' to weave the code results into the report.

Tidy data is a structured way of organizing data where each variable forms a column, each observation forms a row, and each data table stores data about one kind of observation.

This example shows how to replace a slow loop with a faster vectorized operation. The optimized function will run much quicker given that slow_calculation() is vectorized.

This code uses the replicate() function to generate three sets of two normally distributed random numbers. It's similar to running rnorm(2) three separate times and collating the results.