Generate LaTeX code of an estimation result

Convenience functions that generates LaTeX-code of an estimation result in equation-form. printtex can, in principle, be applied to any object for which coef, vcov and logLik methods exist. Note: The generated LaTeX-code contains an eqnarray environment, which requires that the amsmath package is loaded in the preamble of the LaTeX document.

Usage

printtex(x, fitted.name=NULL, xreg.names=NULL, digits=4,
      intercept=TRUE, gof=TRUE, diagnostics=TRUE, nonumber=FALSE,
      nobs="T", index="t", dec=NULL, print.info=TRUE)
  # S3 method for class 'arx'
toLatex(object, ...)
  # S3 method for class 'gets'
toLatex(object, ...)

Arguments

x: an estimation result, e.g. arx, gets or isat object
object: an estimation result of class arx or gets
fitted.name: NULL or a user-specified name of left-hand side variable
xreg.names: NULL or a user-specified character vector with the names of regressors
digits: integer, the number of digits to be printed
intercept: logical or numeric. The argument determines whether one of the regressors is an intercept or not, or its location. If TRUE, then the intercept is assumed to be located at coef(x)[1], and hence the regressor-name of location 1 is excluded from the print. If FALSE, then it is assumed that there is no intercept among the regressors. If numeric, then it is assumed that the regressors contain an intercept at the location equal to the numeric value
gof: logical, whether to include goodness-of-fit in the print
diagnostics: logical, whether to include diagnostics in the print
nonumber: logical, whether to remove or not (default) the equation-numbering
nobs: character, the notation to use to denote the number of observations
index: NULL or a character, only relevant if fitted.name is not NULL, and if the object in question is of class arx, gets or isat
dec: NULL or a character (for example ","). In the latter case, an attempt is made to replace the dot separator . with the character in dec
print.info: logical, whether to print the info at the start or not
...: arguments passed on to printtex

Details

toLatex.arx and toLatex.gets are simply wrappers to printtex

Value

LaTeX code of an estimation result

Author

Genaro Sucarrat, http://www.sucarrat.net/

Examples

##simulate random variates, estimate model:
y <- rnorm(30)
mX <- matrix(rnorm(30*2), 30, 2)
mymod <- arx(y, ar=1:3, mxreg=mX)

##print latex code of estimation result:
printtex(mymod)
#> % Date: Mon Apr  7 09:08:32 2025 
#> % LaTeX code generated in R 4.4.3 by the gets package
#> % Note: The {eqnarray} environment requires the {amsmath} package
#> \begin{eqnarray}
#>   \widehat{y} &=&  - \underset{(0.2278)}{0.0943} + \underset{(0.2377)}{0.3050}ar1 + \underset{(0.2320)}{0.0541}ar2 - \underset{(0.2389)}{0.0103}ar3 - \underset{(0.2714)}{0.2830}mxreg1 - \underset{(0.2562)}{0.1437}mxreg2 \\[2mm] 
#>    && R^2=0.1300 \qquad \widehat{\sigma}=1.1209 \qquad LogL=-38.3939 \qquad T = 27 \nonumber \\ 
#>    && \underset{[p-val]}{ AR(4) }: \underset{[0.7966]}{1.6678}\qquad \underset{[p-val]}{ ARCH(1)}:\underset{[0.1313]}{2.2771}\qquad \underset{[p-val]}{ Normality }:\underset{[0.7758]}{0.5078} \nonumber 
#> \end{eqnarray}

##add intercept, at the end, to regressor matrix:
mX <- cbind(mX,1)
colnames(mX) <- c("xreg1", "xreg2", "intercept")
mymod <- arx(y, mc=FALSE, mxreg=mX)

##set intercept location to 3:
printtex(mymod, intercept=3)
#> % Date: Mon Apr  7 09:08:32 2025 
#> % LaTeX code generated in R 4.4.3 by the gets package
#> % Note: The {eqnarray} environment requires the {amsmath} package
#> \begin{eqnarray}
#>   \widehat{y} &=&  - \underset{(0.2374)}{0.2241}xreg1 - \underset{(0.2024)}{0.1291}xreg2 - \underset{(0.1972)}{0.1021} \\[2mm] 
#>    && R^2=0.0350 \qquad \widehat{\sigma}=1.0692 \qquad LogL=-43.0759 \qquad T = 30 \nonumber \\ 
#>    && \underset{[p-val]}{ AR(1) }: \underset{[0.1043]}{2.6383}\qquad \underset{[p-val]}{ ARCH(1)}:\underset{[0.1111]}{2.5388}\qquad \underset{[p-val]}{ Normality }:\underset{[0.6428]}{0.8839} \nonumber 
#> \end{eqnarray}