periodogram

hana_ml.algorithms.pal.tsa.periodogram.periodogram(data, key=None, endog=None, sampling_rate=None, num_fft=None, freq_range=None, spectrum_type=None, window=None, alpha=None, beta=None, attenuation=None, mode=None, precision=None, r=None)

Periodogram is an estimate of the spectral density of a signal or time series. It can help determine if a particular frequency is a meaningful component of the data or if it is just a random noise.

Parameters:
dataDataFrame

The input data should comprise at least two columns. One is ID column, while the other is raw data.

keystr, optional

The ID column.

Defaults to the first column of data if the index column of data is not provided. Otherwise, defaults to the index column of data.

endogstr, optional

The column of series to be tested.

Defaults to the first non-key column.

sampling_ratefloat, optional

Represents the frequency of sampling in the sequence.

Defaults to 1.0.

num_fftinteger, optional

Denotes the number of DFT (Discrete Fourier Transform) points. If num_fft is smaller than the length of the input data, the input gets trimmed. If it is larger, the input gets appended with zeroes by default.

Defaults to the length of sequence.

freq_range{"one_side", "two_sides"}, optional

Reflects the desired frequency range for the result.

Defaults to "one_side".

spectrum_type{"density", "spectrum"}, optional

Indicates the chosen scale type for the power spectrum.

Defaults to "density".

windowstr, optional

Available input window type:

  • 'none',

  • 'bartlett',

  • 'bartlett_hann',

  • 'blackman',

  • 'blackman_harris',

  • 'bohman',

  • 'chebwin',

  • 'cosine',

  • 'flattop',

  • 'gaussian',

  • 'hamming',

  • 'hann',

  • 'kaiser',

  • 'nuttall',

  • 'parzen',

  • 'tukey'

No default value.

alphafloat, optional

A window parameter for Blackman and Gaussian window. Only valid for Blackman and Gaussian window. Defaults to: - "Blackman" : 0.16. - "Gaussian" : 2.5.

betafloat, optional

A parameter specific to the Kaiser window. Only valid for Kaiser window.

Defaults to 8.6.

attenuationfloat, optional

A parameter specific to the Chebwin window. Only valid for Chebwin window.

Defaults to 50.0.

mode{'symmetric', 'periodic'}, optional

A parameter specific to the Flattop window. Only valid for Flattop window.

Defaults to 'symmetric'.

precision{'none', 'octave'}, optional

A parameter specific to the Flattop window. Only valid for Flattop window.

Defaults to 'none'.

rfloat, optional

A parameter specific to the Tukey window. Only valid for Tukey window.

Defaults to 0.5.

Returns:
DataFrame

Result, structured as follows:

  • ID: ID column.

  • FREQ: Value of sample frequencies.

  • PXX: Power spectral density or power spectrum of input data.

Examples

>>> res = periodogram(data=df,
                      key='ID',
                      endog='X',
                      sampling_rate=100,
                      window="hamming",
                      freq_range="two_sides")
>>> res.collect()