Correlation

What is the Pearson correlation coefficient?

Let (x1, y1),...,(xn, yn) be n points.

Then their Pearson Correlation coefficient is defined as:

∑i=1n(x−x‾)(y−y‾)∑i=1n(x−x‾)2∑i=1n(y−y‾)2\dfrac{\sum_{i=1}^n(x - \overline{x})(y - \overline{y})}{\sqrt{\sum_{i=1}^n(x - \overline{x})^2} \sqrt{\sum_{i=1}^n(y - \overline{y})^2}}

where x‾\overline{x} and y‾\overline{y} are the averages of the x and y values.

This number lies between -1 and 1 and measures their linear correlation.

Examples

Correlation: -0.982

Correlation: -0.603

Correlation: -0.003

Correlation: 0.52

Correlation: 0.998

What does this website do?

This website allows users to guess the correlation of a set of points. The goal is to see how people are at guessing correlations.

When you make a guess, the website collects the actual correlation, your guess and your guess number. The results are available here.

Start guessing! Github