DetrendedFluctuation
DFA exponent: how far a signal remembers its own past.
Explore this node in the cookbook →
data
exponent
- Type name
signal:DetrendedFluctuation- Plane
signal- Tags
analysis- Language
python- Tier
in-process- Bundle
complexity- Source
node-bundles/complexity/detrended_fluctuation.py- Availability
- available
Slots
| Slot | Direction | Type | |
|---|---|---|---|
data | input | ARRAY | |
exponent | output | ARRAY |
Parameters
This node declares no parameters of its own.
Source
The current source of this node, as it stands in the goofi repository at node-bundles/complexity/detrended_fluctuation.py.
"""DetrendedFluctuation — the DFA exponent, from antropy.
Says how a signal's fluctuation grows with the window it is measured over: near 0.5 is noise with
no memory, near 1.0 is long-range correlated, above 1.5 is a random walk. The last axis is time and
is consumed; every axis before it survives.
"""
import antropy
import numpy as np
import goofi
class DetrendedFluctuation(goofi.Node):
"""DFA exponent: how far a signal remembers its own past."""
TAGS = ["analysis"]
INPUTS = {"data": goofi.InputSlot(goofi.DataType.ARRAY, required=True)}
OUTPUTS = {"exponent": goofi.DataType.ARRAY}
def process(self, data):
out = np.apply_along_axis(
antropy.detrended_fluctuation, -1, np.asarray(data.data, dtype=np.float64)
)
return out.astype(np.float32), data.drop_axis(-1)