NESC0825 MESA.
MESA, Fourier Analysis of Maximum Entropy Spectra and Correlation Function Analysis
NAME OR DESIGNATION OF PROGRAM, COMPUTER, DESCRIPTION OF PROBLEM OR FUNCTION, METHOD OF SOLUTION, RESTRICTIONS ON THE COMPLEXITY OF THE PROBLEM, TYPICAL RUNNING TIME, UNUSUAL FEATURES OF THE PROGRAM, RELATED AND AUXILIARY PROGRAMS, STATUS, REFERENCES, MACHINE REQUIREMENTS, LANGUAGE, OPERATING SYSTEM UNDER WHICH PROGRAM IS EXECUTED, OTHER PROGRAMMING OR OPERATING INFORMATION OR RESTRICTIONS, NAME AND ESTABLISHMENT OF AUTHOR, MATERIAL, CATEGORIES
[ top ]
[ top ]
To submit a request, click below on the link of the version you wish to order.
Only liaison officers are authorised to submit online requests. Rules for requesters are
available here.
| Program name | Package id | Status | Status date |
|---|---|---|---|
| MESA | NESC0825/02 | Tested | 14-OCT-1986 |
Machines used:
| Package ID | Orig. computer | Test computer |
|---|---|---|
| NESC0825/02 | IBM 3033 | IBM 3081 |
[ top ]
3. DESCRIPTION OF PROBLEM OR FUNCTION
The MESA program determines periodic components from input data which represent a time series.
Transformations may be performed on the raw data before doing the spectral analysis. Routines are included to calculate both the Fourier spectrum and maximum entropy spectrum, the Yule-Walker estimates of the autocovariance function, the periodogram ordinates, the cumulative periodogram, the autocorrelation function resulting from the maximum entropy computations, and the final prediction error, if desired.
The MESA program determines periodic components from input data which represent a time series.
Transformations may be performed on the raw data before doing the spectral analysis. Routines are included to calculate both the Fourier spectrum and maximum entropy spectrum, the Yule-Walker estimates of the autocovariance function, the periodogram ordinates, the cumulative periodogram, the autocorrelation function resulting from the maximum entropy computations, and the final prediction error, if desired.
[ top ]
4. METHOD OF SOLUTION
A test for white noise is performed using the cumulative periodogram. The Fourier power spectrum and the maximum entropy spectrum are used for estimating the period of the time series. The maximum entropy spectral analysis is based on the spectrum that corresponds to the most random and least predictable time series whose autocorrelation function agrees with the given input values.
A test for white noise is performed using the cumulative periodogram. The Fourier power spectrum and the maximum entropy spectrum are used for estimating the period of the time series. The maximum entropy spectral analysis is based on the spectrum that corresponds to the most random and least predictable time series whose autocorrelation function agrees with the given input values.
[ top ]
[ top ]
6. TYPICAL RUNNING TIME
Approximately 30 seconds are required to process a time series of 100 data points, using 600 frequencies and one maximum entropy spectrum. The time required depends primarily on the length of the time series, the number of frequencies used and the number of times MESA is performed.
Approximately 30 seconds are required to process a time series of 100 data points, using 600 frequencies and one maximum entropy spectrum. The time required depends primarily on the length of the time series, the number of frequencies used and the number of times MESA is performed.
NESC0825/02
NEA-DB compiled, linked, and executed the test case included in this package on an IBM 3081 computer in 30 seconds of CPU time.[ top ]
[ top ]
[ top ]
10. REFERENCES
- B.I. Kirk, B.W. Rust, and W. Van Winkle,
Time Series Analysis by the Maximum Entropy Method,
ORNL-5332, January 1979.
- J.P. Burg,
Maximum Entropy Spectral Analysis,
Ph. D. Thesis, Stanford University, Department of Geophysics,
May 1975.
- B.I. Kirk, B.W. Rust, and W. Van Winkle,
Time Series Analysis by the Maximum Entropy Method,
ORNL-5332, January 1979.
- J.P. Burg,
Maximum Entropy Spectral Analysis,
Ph. D. Thesis, Stanford University, Department of Geophysics,
May 1975.
NESC0825/02, included references:
- B.L. Kirk and B.W. Rust:Inductive Time Series Modeling Program.
ORNL/TM-7534 (October 1985)
- B.L. Kirk, B.W. Rust and W. Van Winkle:
Time Series Analysis by the Maximum Entropy Method
ORNL-5332 (January 1979)
[ top ]
11. MACHINE REQUIREMENTS
250 to 320K bytes of main memory are required with one input unit (logical unit 5) and one output unit (logical unit 6).
250 to 320K bytes of main memory are required with one input unit (logical unit 5) and one output unit (logical unit 6).
NESC0825/02
To execute the test case included in this package on an IBM 3081 computer, 584K bytes of main storage are required.[ top ]
NESC0825/02
MVS/SP (IBM 3081).[ top ]
14. OTHER PROGRAMMING OR OPERATING INFORMATION OR RESTRICTIONS
For
plotting, the program makes use of the Integrated Software Systems Corporation (ISSCO) proprietary software product, DISSPLA. A subroutine with dummy entry points has been supplied by NESC to prevent unresolved external references. The DCADBE and UERTST sub- routines included are from the proprietary International Mathema- tical and Statistical Libraries, Incorporated (IMSL). These routines may not be redistributed or removed for use in other software.
For
plotting, the program makes use of the Integrated Software Systems Corporation (ISSCO) proprietary software product, DISSPLA. A subroutine with dummy entry points has been supplied by NESC to prevent unresolved external references. The DCADBE and UERTST sub- routines included are from the proprietary International Mathema- tical and Statistical Libraries, Incorporated (IMSL). These routines may not be redistributed or removed for use in other software.
[ top ]
[ top ]
NESC0825/02
| File name | File description | Records |
|---|---|---|
| NESC0825_02.001 | This information file | 60 |
| NESC0825_02.002 | MESA program source, FORTRAN 77 | 5739 |
| NESC0825_02.003 | JCL | 59 |
| NESC0825_02.004 | Test case input | 229 |
| NESC0825_02.005 | Test case output | 1974 |
Keywords: statistics.