There are many MacAnova functions that are useful in time series analysis. In addition there are two files, tser.mac and arima.mac, containing macros for doing frequency domain and time domain analyses. Functions useful in time series analysis Fast Fourier transforms (FFT) for Complex, Hermitian, and Real series cft(), hft(), rft() Functions for working with complex and Hermitian series and Fourier transforms cconj(), creal(), cimag(), cpolar(), crect(), cprdc(), cprdcj(), hconj(), hreal(), himag(), hpolar(), hrect(), hprdh(), hprdhj(), cmplx(), ctoh(), htoc(), unwind(), reverse(), padto(), rotate() Autoregressive and moving average operators and their zeros autoreg(), movavg(), polyroot() Other functions convolve(), partacf(), yulewalker() Type, for example, 'usage(cft)' or 'help(cft)', to get a thumbnail sketch or a complete description of cft(). Macros for time series analysis File tser.mac, distributed with MacAnova, contains the following macros: arspectrum Estimate spectrum of autoregression by solving the Yule- Walker equations autocor Compute autocorrelation function autocov Compute autocovariance function burg Estimate autoregression coefficients using Burg's algorithm and optionally compute the spectrum of the fitted moddel compfa Compute smoothed modified periodograms and, optionally, cross periodograms, using cosine tapering, with optional detrending compza Compute the Fourier transform of a cosine tapered series, optionally detrending costaper Compute a cosine taper with a specified amount of tapering crsspectrum Compute smoothed periodgrams and cross periodogram with no tapering or detrending detrend Remove a polynomial trend in equally spaced time dpss Compute discrete prolate spheroidal sequences evalpoly Evaluate a real polynomial of a complex variable ffplot Plot a frequency function against frequency gettsmacros Retrieves macros from tser.mac multitaper Compute multitaper spectrum estimates spectrum Compute smoothed periodograms with no tapering or detrending testnfreq Test whether its argument has prime factors > 29 tsplot Plot time series against time These can be retrieved by, for example, getmacros(multitaper) or multitaper <- macroread("tser.mac","multitaper"). Help for these macros is available in file tser.hlp. You can get help on these macros using help(). If you know a macro is in tser.mac, it may be faster to use pre-defined macro tserhelp(). For example, to get help on burg(), type tserhelp(burg). See topic tserhelp() for details. tserhelp() also can retrieve the following informational topics from tser.hlp: bandwidth Comments about the bandwidth and EDF of spectrum estimates complex_data Information on representing complex data and series in MacAnova complex_fun Summary of MacAnova functions for working with complex series fourier Information concerning Fourier transforms hermitian Information on complex series with Hermitian symmetry See help() for information on direct use of help() to retrieve help information from tser.hlp and arima.mac. File arima.mac, distributed with MacAnova, contains the following macros: acfarma Compute autocovariance function of ARMA model arima Fit ARIMA model or linear regression with ARIMA errors by unconditional least squares or MLE estimation arimares Compute residuals from ARIMA model; used by macros arima, hannriss, innovest. hannriss Fit an ARIMA model using the Hannan-Rissannen algorithm innovations Compute the coefficients for one step prediction in terms of previous one-step prediction errors. Used by macro innovest innovest Fit an ARIMA model using the innovations algorithm levmar Fit a non-linear model by least squares using a form of the Levenberg-Marquart algorithm moveoutroots Fix up coefficients for a MA or AR operator so that all the zeros are outside the unit circle in the complex plane nlreg Fit a non-linear regression model by least squares. rhatcovar Compute variances and or covariances of sample autocorrelations or the entire variance matrix of the sample autocorrelation function using Bartlett's formula rhatvar Compute variances of sample auto correlations using Bartlett's formula specarma Compute spectrum of ARMA model File arima.mac serves as its own help file from which the help topics can be retrived either by help() or by pre-defined macro arimahelp(). If you know a topic is in arima.mac, arimahelp() may be faster since it searches only one file. To get help on, say, macro arima(), type arimahelp(arima) or help(arima). See arimahelp() for details. See help() for information on direct use of help() to retrieve help information from tser.hlp and arima.mac. See also topics 'macros', macroread(), getmacros(), usage(), macrousage().

Gary Oehlert 2003-01-15