     This directory contains test code for the functions of the 'tseries'
     segment of the CCM math library. Each test code file has a sample
     of the correct output appended to it. Input data files named in
     the test code header, are in the 'data' sub-directory.

     The 'data' subdirectory contains time-series model definition files.
     The format of these definition files is specfied below.

     The 'generators' subdirectory contains the source code of programs
     that will generate the sequence of a specfied time series.


     Note: If your machine is big-endian or it employs a floating point
           format different from the IEEE 64 bit standard for doubles
           you must regenerate some of the test data files in the 'data'
           subdirectory. See the 'generators' sub-directory README file
           for directions on this.


     ----------------------------------------
     Format of time series model data files
     ----------------------------------------
 Time series model definition files
-------------------------------------------------------------------------
 A. Ordinary ARMA model definition

    nar  nma           (# autoregressive  # moving average)
    p1   lag1
    -    -           ( parameter - lag list
    -    -              AR then MA , # = nar + nma = np )
    pN   lagN

    random_seed

 B. FACTOR model definition

    nfc nar nma ndif   (# factors # autoregressive # moving average
                        #difference order ) 
    p1    lag1
    -     -            ( parameter - lag list
    -     -              factor , AR , MA ;  # = nfc + nar +nma = np )
    pN    lagN

    random_seed

--------------------------------------------------------------------------

     In the 'data' subdirectory files with a name beginning with 'tsf'
     refer to factor models, while files with a name 'ts' refer to
     ordinary ARIMA models. The sufix describes the type of file, with

          .dat  files contain a model definition
                ( used as input to generate a series)
          .d    files contain a model specification
                ( definition sans parameter values
                  used as input to estimate a time series fit)
          .b    files contain the binary data of a time series
                ( generated by 'garma' (ARIMA) modesl or by
                  'gfarma' (factor models) )
          .er   files contain the resuidal output from a
                model estimation attempt (this is optional
                output from the estimation programs and can
                be used as input to the residual analysis
                program tresid.c)

     Note: It is instructive to run test estimations in which the
           model specification (.d file) does not match the
           definition file used to create the series!

     The following list specifies the command line parameters and
     prompt responses required to run the standard tests, whose
     output is appended to the test source code. A prompt may
     specify several parameters.
 

     tdrfmod    20  data/tfs2.dat
     tdrmod     20  data/ts1.dat
     tevfmod    30  data/tfs0.dat
     tevmod     30  data/ts2.dat
     tfmest     data/tfs0.d  data/tfs0.b
                prompt response sequence: s s s f q n
     tfsread    data/tfs0.b
     tmest      data/ts0.d  data/ts0.b
                prompt response sequence: s s s f f q n
     tparma     data/ts2.dat
     tresid     data/ts0.er
                prompt responses: 20 -4 4 16 : n
     tsany      data/ts0.b
                prompt response: 0 2 20
     tsarma     20  data/ts2.dat
     tsdiff     2  data/ts6.b
     tsread     data/ts0.b
     txmean     data/ts6.b
