How can I transform 3 dimensional data into 1 dimensional data?

I have a bunch of data acquired using an iPhone seismometer app that I need to analayze.

It is presented as (accelerations relative to g), in three dimensions x, y and z. The accelerations oscillate around a baseline close to zero.

I need to be able to somehow integrate these numbers to form a 'tremor intensity' value (which can be dimensionless if necessary), which I could then use to plot a single moving average of the signal's energy.

I am a Matlab newbie.

Answer 1

[Star Strider]

The simplest way may be to take the Euclidean norm of the three channels by row:

S = randi(10, 10, 3);               % Random (10x3) matrix
TI = sqrt(sum(S.^2,2));             % Tremor Intensity

where ‘TI’ is sqrt(x^2+y^2+z^2) for each row of ‘S’