guessTheCorrelation: The game "Guess The Correlation".

Description Usage Arguments Details Value Author(s) References Examples

Description

An R implementation of "Guess the Correlation" - a fun and wholesome game for the whole family. The player is presented with a scatter plot of random data generated from a standard bivariate normal distribution with an unknown correlation coefficient (rho). The objective is to guess rho. Se Details for the rules.

Usage

1
guessTheCorrelation(n=NA, population=FALSE, variation=1)

Arguments

n

The number of random pairs to generate in each round.

If n = NA (default) a random number of random pairs are generated in each round chosen uniformly in seq(50, 250, by=50).

population

If population = TRUE the task is to guess the population correlation from which the random sample was generated.

If population = FALSE (default) the task is to guess the sample correlation coefficient computed from the randomly generated sample.

variation

Not implemented yet.

Details

The player has a number of lives and a certain amount of coins. At the beginning of the game these are initialized to 5 coins and 3 lives.

If the player guesses the correlation coefficient within +/- 0.05 of the true value, the player is awarded an extra life and 5 coins.

If the player guesses the correlation coefficient within +/- 0.1 of the true value, the player is awarded an extra coin.

If the guess exceeds +/- 0.1 the player loses a life.

When there are no lives left it's GAME OVER!

The true correlation coefficients and the guesses are rounded to two decimal places.

Value

A vector of signed differences ([-2;2]) between the guesses and the true correlation coefficients (population or sample depending on the argument).

Author(s)

A. K. Jensen

References

http://guessthecorrelation.com/

Examples

1
## Not run: stats <- guessTheCorrelation(n=100)

aejensen/GTC documentation built on May 10, 2019, 7:28 a.m.