Reserved Names

This lists the variable names reserved for use by RATS. You can also access a categorized list of these variables from within RATS, by selecting the View>Standard Variables operation.

There are two types of reserved variables: accessible variables and special purpose variables.

Accessible variables are set by various RATS instructions. Examples include %RSS (residual sum of squares) and %BETA (vector of coefficients), which are defined by estimation instructions such as LINREG. You can use these variables in expressions just like any other variables. Note that all of these variables begin with the % character.

Special purpose variables serve a particular purpose. These are: the I and J integer variables used as EWISE subscripts; the integer T reserved for use as the time subscript in SET or FRML instructions; the CONSTANT series used for regressions with a constant term; %MVGAVGE, used for adding a moving average lag to an equation; %GARCHV, used for including a variance term in the mean equation in an ARCH or GARCH model; the constant %PI; and the word TO, used for lists on loops, series lists, and in lag notation.

%BETA	VECTOR	Coefficient vector defined by estimation instructions.
%BETASYS	VECTOR	Full coefficient vector, defined by AR1 and others
%CDSTAT	REAL	Test statistic (hypothesis tests, CDF, STATISTICS)
%COEFF	RECTANGULAR	Matrix of equation coefficients defined by ALLOCATE.
CONSTANT	Special	Series name for constant term in regressions.
%CONVERGED	INTEGER	Variable set to 1 if a process converged, and 0 otherwise (non-linear estimations).
%CVCRIT	REAL	Final convergence criterion. Will be set to zero if the subiterations limit was hit (non-linear estimations).
%DOFORPASS	INTEGER	Pass count through DOFOR loop
%DURBIN	REAL	Durbin-Watson statistics. Computed by regressions
%EBW	REAL	Equivalent band width (WINDOW, NPREG, DENSITY).
%EDF	REAL	Equivalent degrees of freedom (WINDOW)
%ERRCODE	INTEGER	The error code of the last error.
%ESALPHA	REAL	Alpha parameter estimated by ESMOOTH.
%ESDELTA	REAL	Delta parameter estimated by ESMOOTH.
%ESGAMMA	REAL	Gamma parameter estimated by ESMOOTH.
%FRACT01	REAL	1st percentile of a series (STATISTICS).
%FRACT05	REAL	5th percentile of a series (STATISTICS).
%FRACT10	REAL	10th percentile of a series (STATISTICS).
%FRACT25	REAL	25th percentile of a series (STATISTICS).
%FRACT75	REAL	75th percentile of a series (STATISTICS).
%FRACT90	REAL	90th percentile of a series (STATISTICS).
%FRACT95	REAL	95th percentile of a series (STATISTICS).
%FRACT99	REAL	99th percentile of a series (STATISTICS).
%FSIGNIF	REAL	Significance level of F statistic (estimation, testing)
%FSTAT	REAL	F statistic (estimation, testing instructions)
%FUNCVAL	REAL	Final function value (estimation instructions)
%GARCHV	SERIES	Within GARCH only, represents the series of variances.
%INTERACTIVE	INTEGER	0,1 indicator for interactive or batch mode
%ITERS	INTEGER	Number of iterations completed
I	INTEGER	I and J are used as index variables for EWISE. They can be used elsewhere, as in a DO loop index.
J	INTEGER
%JBSIGNIF	REAL	Jarque–Bera test significance level (STATISTICS).
%JBSTAT	REAL	Jarque–Bera test statistic (STATISTICS).
%JSIGNIF	REAL	Significance level for %JSTAT.
%JSTAT	REAL	Test statistic for Hansen J-test of overidentifying restrictions (instrumental variables estimations).
%KAPPA	REAL	Correction factor for %EDF and %EBW (TAPER).
%KURTOSIS	REAL	Excess kurtosis of a series (STATISTICS).
%LAGRANGE	VECTOR	VECTOR of Lagrange multipliers (non-linear estimations with constraints).
%LOGDET	REAL	Log determinant of the variance-covariance matrix (ESTIMATE, NLSYSTEM, SUR, VCV).
%LOGL	REAL	Value of the log-likelihood (most estimation).
%MAXENT	INTEGER	Entry containing maximum value (EXTREMUM, CXTREMUM).
%MAXIMUM	REAL	Maximum value of a series (EXTREMUM, STATISTICS, CXTREMUM, TABLE).
%MEAN	REAL	Mean value of a series (STATISTICS, regressions).
%MEANV	VECTOR	Vector of means of the explanatory variables (PRJ)
%MEDIAN	REAL	Median value of a series (STATISTICS).
%MENUCHOICE	INTEGER	Number of the menu item selected in a USERMENU.
%MENUSTATUS	INTEGER	0-1 status variable (0=Cancel) on a MENU/CHOICE.
%MINENT	INTEGER	Entry containing the min. value (EXTREMUM, CXTREMUM).
%MINIMUM	REAL	Minimum value of a series (EXTREMUM, STATISTICS, CXTREMUM).
%MVGAVGE	Special	Special variable name used to add moving average terms to an equation (VADD, GARCH).
%NA	REAL	Code for a missing value.
%NARMA	INTEGER	Number of ARMA terms estimated (BOXJENK).
%NCMOM	INTEGER	Number of variables (CMOMENT,MCOV).
%NCORRECT	INTEGER	Number of cases correct (DDV).
%NDF	INTEGER	Number of degrees of freedom (estimation instructions).
%NDFINDIV	INTEGER	Number of degrees of freedom individual (PSTATS)
%NDFJ	INTEGER	Number of degrees of freedom of J statistic.
%NDFQ	INTEGER	Number of degrees of freedom of Q statistic.
%NDFTEST	INTEGER	Number of degrees of freedom of the test statistic.
%NDFTIME	INTEGER	Number of degrees of freedom (time) (PSTATS)
%NFREE	INTEGER	Number of freely estimated parameters (estimation instructions).
%NGROUP	INTEGER	Number of individuals/groups with valid observations
%NINSTR	INTEGER	Number of instruments (INSTRUMENTS)
%NMISS	INTEGER	Number of skipped/missing observations (estimation instructions; others).
%NOBS	INTEGER	Number of observations (estimation instructions, others).
%NREG	INTEGER	Number of regressors (estimation instructions).
%NREGMEAN	INTEGER	Number of regressors in mean model (GARCH).
%NREGSYSTEM	INTEGER	Number of regressors in entire model (ESTIMATE).
%NVAR	INTEGER	Number of equations estimated (ESTIMATE,GARCH).
%PI	REAL	The constant π.
%PRIOR	RECTANGULAR	The prior matrix for a VAR (SPECIFY, END(SYSTEM))
%PRJCDF	REAL	Predicted probability from ATMEAN or XVECTOR (PRJ)
%PRJDENSITY	REAL	Predicted density from ATMEAN or XVECTOR (PRJ)
%PRJFIT	REAL	Fitted value from ATMEAN or XVECTOR options (PRJ)
%PRJSTDERR	REAL	Prediction standard error, ATMEAN or XVECTOR options (PRJ)
%QSIGNIF	REAL	Significance level of %QSTAT (estimation instructions).
%QSTAT	REAL	Ljung–Box Q-statistic (estimation instructions).
%RBARSQ	REAL	Adjusted R-squared statistic (regression instructions).
%REPORTCOL	INTEGER	Last column filled in a REPORT.
%REPORTROW	INTEGER	Last row filled in a REPORT.
%REPORTTAGCOL	INTEGER	Last column tagged in a REPORT.
%REPORTTAGROW	INTEGER	Last row tagged in a REPORT.
%RESIDS	SERIES	Series of residuals (any univariate estimation with mean model)
%RHO	REAL	First lag autocorrelation coefficient (all single-equation estimation instructions).
%RSQUARED	REAL	Centered R-squared statistic (regressions).
%RSS	REAL	Residual sum of squares (regressions).
%SCALETAP	REAL	Scale factor for spectral estimators (TAPER).
%SEESQ	REAL	Standard error of estimate squared (regressions).
%SHAPE	REAL	Estimated shape parameter for non-normal GARCH.
%SIGMA	SYMMETRIC	Covariance matrix of resids (ESTIMATE, SUR, NLSYSTEM).
%SIGMASQ	REAL	Maximum likelihood estimate of variance (many estimation instructions).
%SIGNIF	REAL	Significance level of %CDSTAT (hypothesis testing instructions, STATISTICS, CDF).
%SKEWNESS	REAL	Skewness of a series (STATISTICS).
%SSERROR	REAL	Sum of squared errors (PSTATS)
%SSINDIV	REAL	Sum of squares (individual) (PSTATS)
%SSTIME	REAL	Sum of squares (individual) (PSTATS)
%STDERRS	VECTOR	Vector of standard errors (estimation instructions).
%SUMLC	REAL	Value of linear combination (SUMMARIZE)
T	INTEGER	Time subscript used in SET and FRML instructions. In general, should not be used for other purposes.
%THEIL	VECT[RECT]	Vector of rectangular arrays containing the forecast statistics produced by the THEIL instruction.
TO	Special	Reserved word used in lists of consecutive integers (e.g., list of lags such as {1 to 4}, or a list of consecutively-numbered series.
%TRSQ	REAL	Number of observations times the raw (uncentered) R-squared statistic (regressions).
%TRSQUARED	REAL	Number of observations times the centered R-squared statistic (regressions).
%TSTATS	VECTOR	Vector of coefficient t-stats (estimation instructions).
%UZWZU	REAL	\({\bf{u'ZWZ'u}}\) for instrumental variables (LINREG, NLLS, NLSYSTEM, SUR).
%VARIANCE	REAL	Sample variance of a series (STATISTICS, regressions, where it's for the dependent variable)
%VARLAGSUMS	RECTANGULAR	Matrix containing the sums of the lag coefficients for a var (ESTIMATE).
%VARLC	REAL	Variance of linear combination of coeffs (SUMMARIZE).
%VECMALPHA	RECTANGULAR	α matrix for error correction model (ESTIMATE).
%VECMPI	RECTANGULAR	\(\Pi\) matrix for error correction model (ESTIMATE).
%VINDIV	REAL	Variance of individual component (PSTATS, PREGRESS).
%VRANDOM	REAL	Variance of random component (PSTATS, PREGRESS).
%VTIME	REAL	Variance of time component (PSTATS, PREGRESS).
%WMATRIX	SYMMETRIC	Final weight matrix for GMM (LINREG, NLLS)
%X	RECTANGULAR	Array of real series defined by ALLOCATE.
%XX	SYMMETRIC	Covariance matrix of coefficients, or for linear regressions (estimation instructions).
%XXSYS	SYMMETRIC	Full covariance matrix (AR1, others)
%Z	CMATRIX	Array of complex series defined by FREQUENCY.