Its latex code is now much readable, still uses dcolumn, still handles user-supplied standard errors, and it adds a host of features like omitting some rows of like controls, and model naming is flexible. Most importantly it matches variable names across models and leaves holes where they belong. Here's a snippet of sample output:
apsrTable(m1,m2,m3,omitcoef=4:17,lev=.1,align="c",model.counter=6,se="robust")
%Latex table generated by Mike Malecki's apsrTable() %Uncomment the following line and the end one to change figure versions %if you are using a full-featured family such as Minion Pro. \figureversion{tabular} \begin{tabular}{cD{.}{.}{-1}D{.}{.}{-1}D{.}{.}{-1}}\hline & \multicolumn{1}{c}{Model 6} & \multicolumn{1}{c}{Model 7} & \multicolumn{1}{c}{Model 8}\\ % & Model 6 & Model 7 & Model 8 \\ (Intercept) & 71.95 & 35.95 & 38.83 \\ & (44.44) & (38.63) & (45.81) \\ maj.parl & -2.03 & -4.81 & -0.86 \\ & (3.38) & (3.52) & (3.98) \\ maj.pres & -10.50 & -14.45 * & \\ & (6.32) & (5.42) & \\ lrCentered & -1.26 & & 3.71 \\ & (5.07) & & (5.52) \\ pro.pres:lrCentered & -3.59 & & -6.27 \\ & (6.61) & & (6.34) \\ maj.parl:lrCentered & -4.46 & & -8.52 \\ & (5.78) & & (6.73) \\ pro.pres & & -8.81 * & -3.36 \\ & & (4.99) & (5.46) \\ N & 46 & 46 & 46 \\ $R^2$ & 0.82 & & 0.81 \\ adj. $R^2$ & 0.70 & & 0.67 \\ AIC & & 310.38 & \\ BIC & & 434.72 & \\ $\log L$ & & -87.19 & \\\hline \multicolumn{ 4 }{l}{ Robust standard errors in parentheses} \\ \end{tabular} \figureversion{proportional}
No comments:
Post a Comment