AOV/TSA 15-SEPT-1992 A. Schwarzenberg-Czerny Purpose: Compute analysis-of-variance periodogramme. The expected value of the periodogramme for pure noise is 1, for uncorrelated observations and 'ncorr', for correlated observations, where 'ncorr' is an average number of correlated observations. The the periodogramme divided by its expected value has Fisher-Snedecor probability distribution F(order-1,nobs/ncorr-1), where 'nobs' is number of observations and 'order' is defined below. For reference see M.N.R.A.S. 241, 153. Syntax: AOV/TSA intab outima start step nsteps [order] [cover] intab = name of input table, it must contain columns :TIME and :VALUE in DOUBLE PRECISION. For numerical reasons it is advisable to subtract mean from :VALUE. outima = name of output image; its first row contains the AOV periodogramme, the second row contains quality flags indicating the quality of phase coverage for a given frequency. Meaning of the flags: 0 - all phases covered, 1 - underfilled phase bins occurred, 2 - empty bins occurred Statistics of bad bins is listed onto terminal. start = start frequency of the periodogramme, in inverse units of :TIME step = frequency step of the periodogramme, in inverse units of :TIME nsteps = number of frequencies of the periodogramme order = number of phase bins to be employed, 1 < order < 101. cover = number of overlapping bin covers to be employed. Each consecutive cover is shifted by 1/(order*cover) in phase. 0 < cover; default cover = 2. See also: SHOW/TSA, DFT/TSA, SCARGLE/TSA Note: By default 'intab', 'outima', 'start', 'step', 'nsteps' and 'order' retain their values from the last use of TSA commands, unless explicitly specified. See SET/TSA for details of this feature. For smooth light curves use low 'order', e.g. 4 or 3 for best sensitivity. For many observations and light curves with sharp features (e.g. pulses, eclipses) use phase bins of width comparable to that of these features. This will boost sensitivity above that attainable with the Scargle method. Note that phase coverage at and near 0 frequency is notoriously poor for any observations. Hence underpopulated or empty bins may occurre. The summary statistics is printed onto the terminal and for details see the outima quality row. Examples: AOV/TSA LCURVE PERIODG 0.01 0.01 100 4 AOV/TSA ? ? 0.2 0.0001 This second command may be used after the first example to inspect the frequency interval from 0.2 till 0.21 at higher resolution.