Rectangular function

The rectangular function is a mathematical tool used to describe a specific type of shape.

It is often called the rect function, the Pi function, or the gate function. Some researchers also refer to it as a unit pulse or a normalized boxcar function. At its simplest, the function is defined by its height and its width. It stays at a constant value for a specific duration and is zero everywhere else. This function is essential for building a rectangular wave. It serves as a fundamental building block in many areas of science and engineering.

To understand how the function works, we must look at its mathematical definition. The function is defined to be 0, 1, or sometimes even undefined at its edges. However, the area under the curve remains the same regardless of these specific edge definitions. A more general version of this shape is the boxcar function. This general version can be centered at any point and can have any duration. It can also be defined using the Heaviside step function. This allows mathematicians to shift the box left or right on a graph.

One of the most interesting things about the rectangular function is what happens during convolution. Convolution is a mathematical process where one function is combined with another. When you convolve a discontinuous rectangular function with itself, the result changes. The first convolution produces a triangular function. This triangular shape is a piecewise defined spline. It is continuous, meaning there are no gaps, but it is not continuously differentiable. This means the corners are still somewhat sharp.

If you perform the convolution process again, the shape becomes even smoother. A second convolution results in a parabolic spline. This shape is both continuous and differentiably continuous. If you convolve it a third time, you get a cubic spline. A fourth convolution produces a fourth-order spline. Each successive convolution creates a pulse that is wider and has a lower maximum height. These pulses become increasingly smooth because their higher-order derivatives become continuous.

The history of this function is tied to the study of information. In 1953, a researcher named Woodward introduced the rect function in his work. His book was titled "Probability and Information Theory, with Applications to Radar." Woodward used the rect function as an ideal cutout operator. He paired it with the sinc function, which acts as an ideal interpolation operator. These two tools work alongside sampling and replicating operators. This work helped explain how signals are processed in radar technology.

The rectangular function also has deep connections to the study of waves and frequencies. The Fourier transform of the rectangular function is the sinc function. This relationship is governed by the Convolution theorem. Because of this, the Fourier transform of a pulse created by multiple convolutions is simply the sinc function raised to a specific power. For example, the Fourier transform of a triangular function is sinc squared. The transform of a parabolic spline is sinc cubed. This allows scientists to predict how different pulse shapes will behave in a frequency spectrum.

In the field of probability, the rectangular function plays another important role. It can be viewed as a probability density function. In this context, it is a special case of the continuous uniform distribution. This means it describes a scenario where outcomes are equally likely within a certain range. Scientists use specific formulas, like the characteristic function and the moment-generating function, to study these distributions. These mathematical descriptions help predict how data is spread out.

Finally, the rectangular function can be used to represent the Dirac delta function. This is done by looking at a limit as the width of the pulse approaches zero. As the pulse becomes extremely narrow and tall, it begins to act like a single, sharp point. The Fourier transform of a Dirac delta function is a constant. This means its frequency spectrum is infinitely broad. As a pulse is shortened in time, it grows larger in its spectrum. This connection helps engineers understand the limits of signal processing.