Back to Volume
Paper: Some IDL routines for Time-Series Analysis
Volume: 61, Astronomical Data Analysis Software and Systems III
Page: 357
Authors: Rusk, Raymond
Abstract: Over the last decade the Electromagnetics Section at DREP has developed an in-house interactive data analysis and graphics software package tailored towards the specific needs of our research group. This software is well suited to most of our data reduction requirements but it is not as flexible nor as extensible, by the end-user, as many of the commercial data analysis and visualization packages which have since become available. Two commercial packages, IDL and MATLAB, have become widely used at DREP because they provide the basic tools needed for numerical simulation, algorithm prototyping and data analysis in one computing environment. Flexible file I/O, convenient graphics and the ability to work with large data sets have attracted the author to IDL. While IDL provides many ready-to-use tools for image processing it has less support for time-series analysis. In the author's work, it is often necessary to estimate instrumental and environmental noise from very long data sequences. An IDL function, patterned after a similar MATLAB routine, to provide the ensemble averaged power spectral density estimate with confidence limits of one or two data sequences, and the cross spectrum between two time series is given. The data windowing function and degree of overlap processing are user selectable. Routines to efficiently display the original time-series and to provide cascade plots of the moving power spectral density estimate are also presented.
Back to Volume