
130 Applying Analytics: A Practical Introduction
FIGURE 6.2
Average daily temperatur e in Nairobi, Kenya, from 1 January 1995 to 24
December 2006.
are correlated knowing the value of one data set gives you some predictive
power over the other. For example, height and weight are correlated – if I
knew a per son’s height, I could guess their weight with more accuracy than if
I had no information. The correlation isn’t p e rfect, however, because there is
a wide range of possible weights for any given height.
Identifying correlations between two data sets is simple, in principle. When
ψ increases, does φ do the same? What about when ψ decreases? The answers ...