pyyeti.cyclecount.sigcount

pyyeti.cyclecount.sigcount(sig, ampbins=10, meanbins=1, right=True, precision=3, retbins=False, use_pandas=True)[source]

Do rainflow cycle counting on a signal.

Parameters:

  • sig (1d array_like) – Signal (vector) to do cycle counting on.

  • ampbins (scalar or 1d array_like) – Defines the cycle amplitude bins; see getbins() for more complete description on how bins are defined.

    ampbins type

    Brief description (see also getbins())

    scalar

    Defines number of bins to use

    vector

    Defines lower bin boundaries

  • meanbins (scalar or 1d array_like) – Defines the cycle mean-value bins; see ampbins description and getbins() for more complete description on how bins are defined.

  • right (bool; optional) – Indicates whether the bins include the rightmost edge or the leftmost edge. If right == True (the default), then the bins [1,2,3,4] indicate (1,2], (2,3], (3,4].

  • precision (integer; optional) – Precision to use for DataFrame labels.

  • retbins (bool; optional) – If True, return the ampbins and meanbins vectors.

  • use_pandas (bool; optional) – If True, this routine will return table as a pandas DataFrame.

Returns:

  • table (pandas DataFrame or 2d ndarray) – A cycle count table (len(meanbins) x len(ampbins)). Each value in table entry is the number of cycles in the bin (see below).

  • ampbins (1d ndarray) – Boundaries of amplitude bins used; length = # of bins + 1. Only returned if retbins is True.

  • meanbins (1d ndarray) – Boundaries of mean-value bins used; length = # of bins + 1. Only returned if retbins is True.

Notes

Steps:

  1. Calls findap() to find all local minima and maxima: peaks = sig[findap(sig)]

  2. Calls rainflow() to do the cycle counting: rf = rainflow(peaks)

  3. Summarizes the rainflow counting by putting the counts in bins. For each mean value bin (i’th):

    1. rf is filtered down to only those rows where the mean value falls in the mean-bin (call this subset of rf rffilt).

    2. For each amplitude bin (j’th): count number of cycles in rffilt with amplitude that falls in the amp-bin and store in table[i, j]

A value falls in bin i if:

bin[i] < value <= bin[i+1]        } if right is True
bin[i] <= value < bin[i+1]        } if right is not True

Examples

>>> from pyyeti import cyclecount
>>> import numpy as np
>>> sig = np.arange(100)
>>> sig[::2] *= -1   # [0, 1, -2, 3, -4, ..., 99]
>>> # `sig` has 99 half-cycles; amplitude grows from 0.5 up to
>>> #  98.5; mean of each is either 0.5 or -0.5
>>> table = cyclecount.sigcount(sig, 2, 2)
>>> table
Amp              (0.402, 49.500]  (49.500, 98.500]
Mean...
(-0.501, 0.000]             12.5              12.0
(0.000, 0.500]              12.5              12.5

We can focus on amplitudes only (ignoring the mean value bins):

>>> table.sum(axis=0)
Amp
(0.402, 49.500]     25.0
(49.500, 98.500]    24.5
dtype: float64

Focus on only the mean value bins:

>>> table.sum(axis=1)
Mean
(-0.501, 0.000]    24.5
(0.000, 0.500]     25.0
dtype: float64

If not using pandas:

>>> cyclecount.sigcount(sig, 2, 2, use_pandas=False)
array([[ 12.5,  12. ],
       [ 12.5,  12.5]])