Next Previous Contents

1. Introduction

During the last decade the number of wavelet applications increased drastically in various scientific fields, in high energy physics (HEP), in particular [1,2]. Wavelet transform (WT) is demanded when one needs simultaneous knowledge of various signal frequencies, and also the time location of these frequencies.

Being a local integral transformation, WT overcomes two main shortcomings of the Fourier transform:

• Being nonlocal and having infinite support $(-\infty,\infty)$, Fourier transform requires a broad band of frequencies to decompose even a short signal;
• The distortion of a single harmonics in Fourier space affects, after reconstruction, the whole signal/image in coordinate space.

The main idea of the wavelet transform is to decompose the function under consideration with respect to a functional basis built by dilations and shifts of a single well localized function, called a basic wavelet.

For effective computer implementation of wavelet algorithms, the shifts and dilations should be in mutual concord. Fortunately, for a wide class of functions, used as basic wavelets, such algorithms exist. They have a pyramidal structure reducing by factor 2 the number of degrees of freedom being calculated at each step. These algorithms are usually referred to as fast wavelet transform (FWT).

Accomplishing of a wavelet analysis usually means the following:

1. make a proper choice of a wavelet type;
2. fulfill a wavelet filtering for denoising, removing pedestals and extracting some features of analyzed data. It is carried out by
   1. transforming data to the wavelet domain;
   2. applying desirable cuts on wavelet 2D-spectrum;
   3. making inverse transform.

Wavelet applications to analyze 2D-patterns in the phase space of nuclear-nuclear events with the high multiplicity allows to reveal or emphasize some local properties of individual events inherent, say, for jets, as correlations or the presence of a dense group of particles.

In this paper after a long write up of one- and two-dimensional wavelet formalism and computational details we introduce a new version of the WASP (Wavelet Analysis of Secondary Particles angular distributions) package. The first version was described in [3] and successfully applied for data analysis [4].

Let us remind that WASP is a C++ program aimed to analyze angular distributions¹ of secondary particles generated in nuclear interactions. (It is designed for data analysis of the STAR and ALICE experiments.) It uses a wavelet analysis for this purpose and the vanishing momenta wavelets are chosen as a basis [5]. WASP version 1.2 allows one to perform both one- and two-dimensional wavelet analysis.

Next Previous Contents