Functional Data Analysis - The Basics of Basis Functions
Functional Data Analysis - The Basics of Basis Functions
-What is a Basis
Function Expertise: Beginner
-How do I know
what basis system to
use? What is a basis function?
Here's the problem. We want a mathematical description of a curve or any other data
The distributed over space, time, and other types of continuum. Flexibility is a central issue
Characteristics of
Spline Functions since we usually cannot say in advance how complex the curve will be, or specify certain
The of its characteristics. We don't have the time or patience to search some handbook of
Characteristics of
Spline Functions known functions for one that looks like what we want to study, either. Moreover,
II however it is that we design our mathematical function, we will want to do any
computation that is needed to fit the data quickly and with a minimum of programming.
Goods
We need, therefore, a set of basic functional building blocks that can be stacked on top
of one another so as to have the features that we need. Mathematical Lego, Meccano,
Dissecting the
U.S. Nondurable Erector, Tinker Toy, or whatever, in other words. Since this is mathematics, we use the
Goods Index symbol øk(t) to stand for the kth function in our toy box, and we call this the kth basis
Smoothing the
Nondurable function.
Goods Index
Phase-Plane By "stacking" in mathematics we mean adding things, possibly after multiplying each of
Plotting of the
Goods Index them by its own constant. So here is how we will construct a function f(t) using K of
these blocks:
The Features of In math-speak, this is a "linear combination". The construction of an actual function
Human Growth
Data then becomes a matter of assigning values to the Κ constants ak.
From Raw Data to
Functional Data What are some examples of commonly used basis functions?
Smoothing the There are many ways to put together a kit of basis functions. The oldest system consists
Growth Data
of the powers of t, that is, 1, t, t2, ... , tk. Linear combinations of these functions are
Monotone
Smoothing of the called polynomials.
Growth Data
A system developed in early nineteenth century France is the Fourier basis, consisting of
Weather 1 and a series of pairs of sines and cosines of increasing frequency; that is, 1, sin(ωt),
cos(ωt), sin(2ωt), cos(2ωt), sin(3ωt), cos(3ωt),.... The constant ω plays an important
role: functions constructed with this system repeat themselves each time t increases by
Weather
Introduction 2π/ω units. Consequently, we tend to use this basis for periodic functions. Figure 1
Summarizing the shows the first five basis functions in the series with a period of one unit.
Functional Data
Weather Data
Analysis
Figure 1: The first five functions in the Fourier series with a period of one unit.
Polynomials, though simple and familiar, are actually not that flexible, and consequently
they have been more or less replaced by a basis system called the spline basis system.
Splines, although related to polynomials systems, require some further explanation, and
we have some special notes on them in this website. However, you can have a
preliminary look at spline basis functions in figure 2. There are plenty of other basis
systems around, and it is worth mentioning wavelets at this point, without ordering any
description of this system in these introductory notes.
Figure 2: The eight B-spline basis functions of order four defined by the breakpoints shown as dotted lines.