Biology Reference
In-Depth Information
1200
1200
1100
1100
1000
1000
900
900
800
800
700
700
600
600
500
500
400
400
300
300
200
200
100
100
0
0
−
100
−
100
0
24
48
72
96 120 144 168 192 216
240
0
24
48
72
96 120 144 168 192 216
240
Time (hours)
Time (hours)
80
40
0
ARFILTER Analysis of
Noisy Non-Stationary Cosine Wave
−
40
−
80
0
24
48
72
96 120 144 168 192 216
240
Time (hours)
FIGURE 11-17.
Example of ARFILTER applied to a noisy nonstationary cosine wave.
question here is whether the rhythmic characteristics of the filtered
data remain the same as, or at least reasonably close to, those present
in the original unfiltered data. We consider such comparisons in
some of the rhythm analyses examples presented in the following
sections.
B. Detrending
The removal of a slow gradual change (or drift) from the time series is
called detrending. The presence of trend reflects the change in some
quantity or property, such as the gradual depletion of luciferin that
causes a decline in the average levels of bioluminescence. As noted in
the previous section, detrending may be viewed as a special type of
filtering for which the goal is to remove the lowest frequencies. The
residual data are then used for analyses, because, for these data, the low
frequencies (that is, the trend) have been removed.
In this chapter, we use a detrending program called DTRNDANL, which
implements a trend-removing algorithm that requires, among other
things, that users specify the following inputs:
