# Solution 4.5 - Chemometrics: Data Analysis for the Laboratory and Chemical Plant

## Education Article

**Published:**Jan 1, 2000**Channels:**Chemometrics & Informatics

1. The standardised data are given as follows.

Notice that it is normal to use the population rather than sample standard deviation.

2. The scores are as follows.

The first two eigenvalues are 201.33 and 6.32. The sum of squares of the standardised data is 30 ´ 7 = 210, hence the percentage variance is 207.65 / 210 = 98.81%.

3. The scores plot is presented below, with different symbols used to indicate the two groups.

There is good discrimination between the two groups, although primarily along PC1.

4.

a. The scaled data are as follows.

b. The predicted scores are as follows.

c. Superimposing on the scores plot demonstrates that the unknown is 2035.

5. This is because the loadings matrix is orthonormal so that the matrix (** P.P'**) is a unit matrix.