Functional series have a long history that can be traced back to the
early nineteenth century. French mathematician (and politician)
Jean-Baptiste-Joseph Fourier, decomposed a continuous, periodic on
function
into the series od sines and cosines,
![]() |
![]() |
|
![]() |
There are three types of Fourier transforms: integral, serial, and discrete. Next, we focus on discrete transforms and some modifications of the integral transform.
The discrete Fourier transform (DFT) of a sequence
is defined as
![]() |
![]() |
The DFT can be interpreted as the multiplication of the input vector
by a matrix; therefore, the discrete Fourier transform is a linear
operator. If
, then
. The matrix
is unitary (up
to a scale factor), i.e.,
, where
is the
identity matrix and
is the conjugate transpose of
.
There are many uses of discrete Fourier transform in statistics. It
turns cyclic convolutions into component-wise multiplication, and the
fast version of DFT has a low computational complexity of
, meaning that the number of operations needed to transform
an input of size
is proportional to
. For a theory and
various other uses of DFT in various fields reader is directed to
Brigham (1988).[5]
We focus on estimation of a spectral density from the observed data, as an important statistical task in a variety of applied fields in which the information about frequency behavior of the phenomena is of interest.
Let
be a a real, weakly stationary time series
with zero mean and autocovariance function
. An absolutely summable complex-valued function
defined on integers is the autocovariance function of
if and
only if the function
![]() |
A traditional statistic used as an estimator of the spectral density
is the periodogram.
The periodogram , based on a sample
is defined as
![]() |
(7.5) |
Calculationally, the periodogram is found by using fast Fourier transform. A simple matlab m-function calculating the periodogram is
function out = periodogram(ts) out = abs(fftshift(fft(ts - mean(ts)))).^2/(2*pi*length(ts));
An application of spectral and log-spectral estimation involves famous Wolf's sunspot data set. Although in this situation the statistician does not know the ''true'' signal, the theory developed by solar scientists helps to evaluate performance of the algorithm.
The Sun's activity peaks every years, creating storms on the
surface of our star that disrupt the Earth's magnetic field. These
''solar hurricanes'' can cause severe problems for electricity
transmission systems. An example of influence of such periodic
activity to everyday life is 1989 power blackout in the American
northeast.
Efforts to monitor the amount and variation of the Sun's activity by
counting spots on it have a long and rich history. Relatively complete
visual estimates of daily activity date back to 1818, monthly averages
can be extrapolated back to 1749, and estimates of annual values can
be similarly determined back to 1700. Although Galileo made
observations of sunspot numbers in the early 17th century, the modern
era of sunspot counting began in the mid-1800s with the research of
Bern Observatory director Rudolph Wolf, who introduced what he called
the Universal Sunspot Number as an estimate of the solar
activity. The square root of Wolf's yearly sunspot numbers are given
in Fig. 7.4a, data from Tong (1996)[22], p. 471.
Because of wavelet data processing we selected a sample of size
a power of two, i.e., only observations from 1733 till 1998. The
square root transformation was applied to symmetrize and de-trend the
Wolf's counts. Figure 7.4b gives a raw periodogram, while
Fig. 7.4c shows the estimator of log-spectral density
(Pensky and Vidakovic, 2003).[21]
![]() ![]() ![]() |
The estimator reveals a peak at frequency
, corresponding to the Schwabe's cycle ranging from
to
(years), with an average of
years. The Schwabe cycle is the period between two
subsequent maxima or minima the solar activity, although the solar
physicists often think in terms of a
-year magnetic cycle since
the sun's magnetic poles reverse direction every
years.
Windowed Fourier Transforms are important in providing simultaneous insight in time and frequency behavior of the functions. Standard Fourier Transforms describing the data in the ''Fourier domain'' are precise in frequency, but not in time. Small changes in the signal (data) at one location cause change in the Fourier domain globally. It was of interest to have transformed domains that are simultaneously precise in both time and frequency domains. Unfortunately, the precision of such an insight is limited by the Heisenberg's Uncertainty Principle.
Suppose is a signal of finite energy. In mathematical terms,
the integral of its modulus squared is finite, or shortly,
belongs
to
space.
The integral Fourier transform of the signal
Windowed Fourier transform (also called short time Fourier transform, STFT) was introduced by Gabor (1946)[12], to measure time-localized frequencies of sound. An atom in Gabor's decomposition is defined via:
![]() |
If
, then windowed Fourier transform is defined as
The spectrogram,
![]() |
The following are some basic properties of STFT. Let
. Then
![]() |
(7.8) |
![]() |
(7.9) |
Let
. There exist
such
that
if and only if
![]() |
(7.10) |
We next describe the Hilbert transform and its use in defining instantaneous frequency, an important measure in statistical analysis of signals.
The Hilbert transform of the function signal is defined by
Because of the possible singularity at , the integral is to be
considered as a Cauchy principal value, (VP). From (7.12)
we see that
is a convolution,
.
The spectrum of is related to that of
. From the
convolution equation,
![]() |
In statistical signal analysis this associated complex function
is known as analytic signal (or causal signal, since
, for
). Analytic signals are important since they possess
unique phase
which leads to the definition of the
instantaneous frequency.
If is represented as
, then
the quantity
is instantaneous frequency of the
signal
, at time
. For more discussion and use of
instantaneous frequency, the reader is directed to Flandrin (1992,
1999).[10,11]
Wigner-Ville Transform (or Distribution) is the method to represent data (signals) in the time/frequency domain. In statistics, Wigner-Ville transform provide a tool to define localized spectral density for the nonstationary processes.
Ville (1948)[24] introduced the quadratic form that measures a local time-frequency energy:
![]() |
The Wigner-Ville transform is always real since
has Hermitian symmetry
in
.
Time and frequency are symmetric in
, by applying
Parseval formula
one gets,
For any
Integral (7.13) states that one-dimensional Fourier transform
of
, with respect to
is,
![]() |
For example,
![]() |
![]() |
Next we show that expected value of Wigner-Ville transform of
a random process can serve as a definition for generalized spectrum of
a non-stationary process. Let be real-valued zero-mean random
process with covariance function
![]() |
Now, if the process is stationary, then
is
a function of
only and
![]() |
For arbitrary process Flandrin (1999)[11] defined ''power spectrum'' as
![]() |
![]() |
For more information on Wigner-Ville transforms and their statistical use the reader is directed to Baraniuk (1994), Carmona, Hwang and Torresani (1998) Flandrin (1999), and Mallat (1999)[3][6][11][18], among others.