p

,

when the paste water content increased.

he visualisation of fitted.values between the oil content and stone weight

e oil content data. The dots are the raw data and the solid line represents the

egression function.

m function also records the prediction errors which were saved in

e called residuals of the lm object. Figure 4.12(a) shows the

n error distribution and Figure 4.12(b) shows the QQ-plot for the

sed residues (errors). Both the histogram and the QQ-plot

ate that the residues do follow a Gaussian distribution. The

sed errors are defined as below,

ߝെߤఌ

ߪఌ

(4.22)

(a) (b)

The distribution of the residues for the linear regression model constructed

e olive oil content and the paste water. (a) The histogram. (b) The QQ-plot.