# Vertical table lines are discontinuous with booktabs

I created a table in LaTeX, but there are some problems with the borders. The borders remain open whereas I want them closed. Also, in the last row, I want the vertical line gone, ie I want just one column in the last row. Can someone please fix it? Here are my code and my table:

\documentclass{article}
\usepackage{booktabs}

\begin{document}

\begin{table} [h!]
\begin{center}
\begin{tabular}{|l|c|c|c|c|}
\hline
Parameters  & \multicolumn{2}{c}{Model 1} & \multicolumn{2}{c}{Model 2}\\
\midrule

&  Coefficient & 95\% CI
&   Coefficient & 95\% CI \\
\midrule

$\beta_{concern_2}$  &   $0.190^{\ast}$ & (0.113, 0.268) & 0.171 & (0.100, 0.241)\\
$\beta_{concern_3}$   &  0.117 &  (0.043, 0.191)  & 0.117 & (0.050, 0.183)\\
$\beta_{concern_4}$ & 0.210  &  (0.139, 0.281)  & 0.190 & (0.127, 0.253)\\
$\beta_{concern_5}$ & 0.204 & (0.135, 0.273)   & 0.111 & ( 0.049, 0.173)\\
$\beta_{breath_2}$  &  0.157 & (0.07,8 0.236) & 0.208 & (0.136, 0.280) \\
$\beta_{breath_3}$ & 0.115 & (0.041, 0.189) & 0.100 & (0.034, 0.166) \\
$\beta_{breath_4}$ & 0.236 &  (0.160, 0.311) & 0.301 & ( 0.234, 0.368)\\
$\beta_{breath_5}$ & 0.092 & (0.020, 0.163) & 0.079 & (0.015, 0.144) \\
$\beta_{weath_2}$  &   0.164 & ( 0.092, 0.236) & 0.137 & (0.071, 0.203) \\
$\beta_{weath_3}$  & 0.160 &  (0.089, 0.231) & 0.199 & (0.135, 0.263) \\
$\beta_{weath_4}$ &  0.141 & (0.067, 0.215) & 0.133 & (0.066, 0.199) \\
$\beta_{weath_5}$ & 0.176 & (0.103, 0.249) &  0.257 & (0.191, 0.323)\\
$\beta_{sleep_2}$  & 0.111 & (0.036, 0.187) & 0.135 & (0.068 0.203)  \\
$\beta_{sleep_3}$  & 0.110 & (0.036 0.184) & 0.176 &  0.110 0.242 \\
$\beta_{sleep_4}$  &  0.131 & (0.056 0.205 & 0.162 & 0.095 0.229 \\
$\beta_{sleep_5}$  & 0.011 & (-0.064, 0.086) & 0.034 & (-0.033, 0.101) \\
$\beta_{act_2}$ &  0.135 & (0.060, 0.209) & 0.033 & (-0.033, 0.100) \\
$\beta_{act_3}$ &0.195 & (0.121, 0.269) & 0.203 & ( 0.137, 0.268)\\
$\beta_{act_4}$ & 0.214 & ( 0.139, 0.290) & 0.254 & (0.186 0.321) \\
$\beta_{act_5}$ & 0.224 & (0.154, 0.294) & 0.158 & (0.095, 0.221) \\
\hline
& *statistically significant at 5\% level \\
\hline
\end{tabular}
\caption{Regression Coefficients of model 1 and model 2 }
\label{beta}
\end{center}
\end{table}

\end{document}


• Sorry, but shouldn't you add a | also in the \multicolums (i.e: \multicolumn{2}{c|}{Model 1} )? – Claudio Fiandrino Jan 2 '13 at 10:57
• The commands provided by booktabs work correctly only in absence of vertical rules. Remove them and your table will be better (use also \toprule and \bottomrule, though). – egreg Jan 2 '13 at 11:07
• As the Indices are words you could do the follwing $\beta_{\text{concern}_2}$, to do that just add \usepackage{amsmath} to your preamble. – rtzll Jan 2 '13 at 11:20
• This is not really a minimum working example. BTW did you read the booktabs documentation? It discourages \hline and |. Perhaps that's why the output looks so bad. – user10274 Jan 2 '13 at 11:36

That's by design: the author of booktabs hates vertical rules in tables and I fully agree with him. You could act on spacing parameters, namely

• \abovetopsep (0pt by default), used above a \toprule
• \belowbottomsep (0pt by default), used below a \bottomrule
• \aboverulesep (0.4ex by default), used above a \midrule, \cmidrule or \bottomrule
• \belowrulesep (0.65ex by default), used below a \midrule, \cmidrule or \toprule

They are all rigid length (no plus or minus specifications are allowed and they wouldn't make sense anyway).

By (locally) setting these parameters to zero, the vertical rules will match, but it would simpler not to use booktabs commands at all: the heavier \toprule and \bottomrule would be completely out of place.

Don't use vertical rules and the appearance of your table will improve immediately.

\documentclass{article}
\usepackage{amsmath}
\usepackage{booktabs}
\usepackage{array}
\newcolumntype{L}{>{$}l<{$}}
\newcolumntype{C}{>{$}c<{$}}
\newcolumntype{R}{>{$}r<{$}}
\newcommand{\nm}[1]{\textnormal{#1}}

\begin{document}

\begin{table} [h!]
\centering
\begin{tabular}{LCRCR}
\toprule
\multicolumn{1}{l}{Parameters} &
\multicolumn{2}{c}{Model 1}    &
\multicolumn{2}{c}{Model 2}    \\
\cmidrule(lr){2-3}
\cmidrule(lr){4-5}

&
\multicolumn{1}{c}{Coefficient} &
\multicolumn{1}{c}{95\% CI}     &
\multicolumn{1}{c}{Coefficient} &
\multicolumn{1}{c}{95\% CI}     \\
\midrule

\beta_{\nm{concern}_2} & 0.190\makebox[0pt][l]{$^{\ast}$}
& ( 0.113, 0.268) & 0.171 & ( 0.100, 0.241) \\
\beta_{\nm{concern}_3} & 0.117 & ( 0.043, 0.191) & 0.117 & ( 0.050, 0.183) \\
\beta_{\nm{concern}_4} & 0.210 & ( 0.139, 0.281) & 0.190 & ( 0.127, 0.253) \\
\beta_{\nm{concern}_5} & 0.204 & ( 0.135, 0.273) & 0.111 & ( 0.049, 0.173) \\
\beta_{\nm{breath}_2}  & 0.157 & ( 0.078, 0.236) & 0.208 & ( 0.136, 0.280) \\
\beta_{\nm{breath}_3}  & 0.115 & ( 0.041, 0.189) & 0.100 & ( 0.034, 0.166) \\
\beta_{\nm{breath}_4}  & 0.236 & ( 0.160, 0.311) & 0.301 & ( 0.234, 0.368) \\
\beta_{\nm{breath}_5}  & 0.092 & ( 0.020, 0.163) & 0.079 & ( 0.015, 0.144) \\
\beta_{\nm{weath}_2}   & 0.164 & ( 0.092, 0.236) & 0.137 & ( 0.071, 0.203) \\
\beta_{\nm{weath}_3}   & 0.160 & ( 0.089, 0.231) & 0.199 & ( 0.135, 0.263) \\
\beta_{\nm{weath}_4}   & 0.141 & ( 0.067, 0.215) & 0.133 & ( 0.066, 0.199) \\
\beta_{\nm{weath}_5}   & 0.176 & ( 0.103, 0.249) & 0.257 & ( 0.191, 0.323) \\
\beta_{\nm{sleep}_2}   & 0.111 & ( 0.036, 0.187) & 0.135 & ( 0.068, 0.203) \\
\beta_{\nm{sleep}_3}   & 0.110 & ( 0.036, 0.184) & 0.176 & ( 0.110, 0.242) \\
\beta_{\nm{sleep}_4}   & 0.131 & ( 0.056, 0.205) & 0.162 & ( 0.095, 0.229) \\
\beta_{\nm{sleep}_5}   & 0.011 & (-0.064, 0.086) & 0.034 & (-0.033, 0.101) \\
\beta_{\nm{act}_2}     & 0.135 & ( 0.060, 0.209) & 0.033 & (-0.033, 0.100) \\
\beta_{\nm{act}_3}     & 0.195 & ( 0.121, 0.269) & 0.203 & ( 0.137, 0.268) \\
\beta_{\nm{act}_4}     & 0.214 & ( 0.139, 0.290) & 0.254 & ( 0.186, 0.321) \\
\beta_{\nm{act}_5}     & 0.224 & ( 0.154, 0.294) & 0.158 & ( 0.095, 0.221) \\
\midrule[\heavyrulewidth]
\multicolumn{5}{l}{\footnotesize$^*$ statistically significant at 5\% level} \\
\bottomrule
\end{tabular}
\caption{Regression Coefficients of model 1 and model 2}\label{beta}
\end{table}

\end{document}


1. The subscripts are upright, being words

2. The alignment is improved by using features of the table itself; for instance, the third and fifth column are right aligned because of the minus signs only in the first coordinate; it wouldn't be so if the minus sign appeared also in the second coordinate or the headers had been wider.

3. With \cmidrule it's easier to show how the headers group the columns.

4. A trick is used for avoiding the * to take up space.

5. An array trick is used for setting all columns in math mode, ensuring that the minus signs are printed as such.

6. \centering is used instead of the center environment (that adds vertical space).

If you need to change "95% CI" to "95% Bayesian Interval", the best is to split the long phrase into two lines: modify the block

\multicolumn{1}{c}{Coefficient} &
\multicolumn{1}{c}{95\% CI}     &
\multicolumn{1}{c}{Coefficient} &
\multicolumn{1}{c}{95\% CI}     \\
\midrule


into

\multicolumn{1}{c}{Coefficient} & \multicolumn{1}{c}{95\% Bayesian} &
\multicolumn{1}{c}{Coefficient} & \multicolumn{1}{c}{95\% Bayesian} \\
& Interval                          &
& Interval                          \\
\midrule

• I would add a thin \midrule between each group of 4 lines, to increase the legibility of the table. – yo' Jan 2 '13 at 12:39
• This does not answer the OP's question (nor mine). – Brandon Kuczenski Jan 23 '16 at 6:23
• Then OP asked the wrong question :) – Brandon Kuczenski Jan 23 '16 at 20:06
• I'd still really like to know how to fix the vertical line issue instead : / (Can you add a "If you really must do it." section? That would be great!) For me some lines don't hurt to separate content of tables. It sort of seems like many people are only justifying a lack of an easy way of doing that by saying "it looks better without". However, that's subjective. – Zelphir Kaltstahl Mar 11 '16 at 21:33
• @Zelphir Don't use the booktabs commands, but \hline – egreg Mar 11 '16 at 21:59

I don't think the answer "Don't use vertical rules" is very constructive when the question is about making vertical lines look nice. So here is a solution if you only want to modify the width of horizontal lines and if you are willing to not use booktabs:

\documentclass{article}
\usepackage{array}

\begin{document}

\begin{table} [h!]
\centering
\begin{tabular}{|l|c|c|c|c|}
\hline
Parameters  & \multicolumn{2}{c}{Model 1} & \multicolumn{2}{c}{Model 2}\\
\noalign{\hrule height 1.5pt}

&  Coefficient & 95\% CI
&   Coefficient & 95\% CI \\
\noalign{\hrule height 1.5pt}

$\beta_{concern_2}$  &   $0.190^{\ast}$ & (0.113, 0.268) & 0.171 & (0.100, 0.241)\\
$\beta_{concern_3}$   &  0.117 &  (0.043, 0.191)  & 0.117 & (0.050, 0.183)\\
$\beta_{concern_4}$ & 0.210  &  (0.139, 0.281)  & 0.190 & (0.127, 0.253)\\
$\beta_{concern_5}$ & 0.204 & (0.135, 0.273)   & 0.111 & ( 0.049, 0.173)\\
$\beta_{breath_2}$  &  0.157 & (0.07,8 0.236) & 0.208 & (0.136, 0.280) \\
$\beta_{breath_3}$ & 0.115 & (0.041, 0.189) & 0.100 & (0.034, 0.166) \\
$\beta_{breath_4}$ & 0.236 &  (0.160, 0.311) & 0.301 & ( 0.234, 0.368)\\
$\beta_{breath_5}$ & 0.092 & (0.020, 0.163) & 0.079 & (0.015, 0.144) \\
$\beta_{weath_2}$  &   0.164 & ( 0.092, 0.236) & 0.137 & (0.071, 0.203) \\
$\beta_{weath_3}$  & 0.160 &  (0.089, 0.231) & 0.199 & (0.135, 0.263) \\
$\beta_{weath_4}$ &  0.141 & (0.067, 0.215) & 0.133 & (0.066, 0.199) \\
$\beta_{weath_5}$ & 0.176 & (0.103, 0.249) &  0.257 & (0.191, 0.323)\\
$\beta_{sleep_2}$  & 0.111 & (0.036, 0.187) & 0.135 & (0.068 0.203)  \\
$\beta_{sleep_3}$  & 0.110 & (0.036 0.184) & 0.176 &  0.110 0.242 \\
$\beta_{sleep_4}$  &  0.131 & (0.056 0.205 & 0.162 & 0.095 0.229 \\
$\beta_{sleep_5}$  & 0.011 & (-0.064, 0.086) & 0.034 & (-0.033, 0.101) \\
$\beta_{act_2}$ &  0.135 & (0.060, 0.209) & 0.033 & (-0.033, 0.100) \\
$\beta_{act_3}$ &0.195 & (0.121, 0.269) & 0.203 & ( 0.137, 0.268)\\
$\beta_{act_4}$ & 0.214 & ( 0.139, 0.290) & 0.254 & (0.186 0.321) \\
$\beta_{act_5}$ & 0.224 & (0.154, 0.294) & 0.158 & (0.095, 0.221) \\

\hline
& *statistically significant at 5\% level \\
\hline
\end{tabular}
\caption{Regression Coefficients of model 1 and model 2 }
\label{beta}
\end{table}

\end{document}


\midrule , \toprule , and \bottomrule are replaced by \noalign{\hrule height 1.5pt} with the width that you want.

You can also set the width of vertical lines by using !{\vrule width 2pt} instead of | in the tabular parameters.

I understand that this solution is not optimal, but it can do the trick and I hope it would help.

• Welcome to TeX.SX! Don't use the center environment in a table, but the \centering declaration (which doesn't add unwanted vertical space at top and bottom). – egreg Dec 31 '14 at 16:41

As @Puck says in his answer, and as Brandon Kuczenski comments, the responses here are not constructive, neither the question is answered.

The way to use vertical lines with booktabs (despite it is discouraged), is to reduce the space above and below the midrule:

 \aboverulesep=0ex
\belowrulesep=0ex


This will cramp your tables a little bit, so, to return to a regular table line height, you should stretch your array with:

\renewcommand{\arraystretch}{SF}


Where SF is the stretch factor (use SF>1.0).

Just in case, you should add those lines in the preamble.

• But using this solution makes using booktabs very pointless. – Johannes_B Nov 5 '17 at 10:41
• But it answers the question ;) – user215982 Nov 6 '17 at 8:40
• And, BTW, it's not pointless as you can use \begin{tabular}{c|c} together with \cmidrule{2-2} if that's the way you want to do it. – user215982 Nov 6 '17 at 8:45

booktabs also offers the specialrule like so:

\specialrule{\cmidrulewidth}{0pt}{0pt}


which will produce a horizontal line with no spacing and the vertical lines will cross it neatly.

I trimmed it a little, added small spaces between every 4th row, and used short forms for the titles, and right aligned the first column, to tighten it.

\documentclass{article}
\usepackage{amsmath}
\usepackage{booktabs}
\usepackage{array}
\newcolumntype{L}{>{$}l<{$}}
\newcolumntype{C}{>{$}c<{$}}
\newcolumntype{R}{>{$}r<{$}}
\newcommand{\nm}[1]{\textnormal{#1}}

\begin{document}

\begin{table} [h!]
\centering
\begin{tabular}{@{}RCRCR@{}}
\toprule
%\multicolumn{1}{r}{Parameter}
&
\multicolumn{2}{c}{Model 1}    &
\multicolumn{2}{c}{Model 2}    \\
\cmidrule(lr){2-3}
\cmidrule(lr){4-5}
\multicolumn{1}{r}{Param}
&
\multicolumn{1}{c}{Coeff} &
\multicolumn{1}{c}{95\% CI}     &
\multicolumn{1}{c}{Coeff} &
\multicolumn{1}{c}{95\% CI}     \\
\midrule
\beta_{\nm{concern}_2} & 0.190\makebox[0pt][l]{$^{\ast}$}
& ( 0.113, 0.268) & 0.171 & ( 0.100, 0.241) \\
\beta_{\nm{concern}_3} & 0.117 & ( 0.043, 0.191) & 0.117 & ( 0.050,0.183) \\
\beta_{\nm{concern}_4} & 0.210 & ( 0.139, 0.281) & 0.190 & ( 0.127, 0.253) \\
\beta_{\nm{concern}_5} & 0.204 & ( 0.135, 0.273) & 0.111 & ( 0.049, 0.173) \\[4pt]
\beta_{\nm{breath}_2}  & 0.157 & ( 0.078, 0.236) & 0.208 & ( 0.136, 0.280) \\
\beta_{\nm{breath}_3}  & 0.115 & ( 0.041, 0.189) & 0.100 & ( 0.034, 0.166) \\
\beta_{\nm{breath}_4}  & 0.236 & ( 0.160, 0.311) & 0.301 & ( 0.234, 0.368) \\
\beta_{\nm{breath}_5}  & 0.092 & ( 0.020, 0.163) & 0.079 & ( 0.015, 0.144) \\[4pt]
\beta_{\nm{weath}_2}   & 0.164 & ( 0.092, 0.236) & 0.137 & ( 0.071, 0.203) \\
\beta_{\nm{weath}_3}   & 0.160 & ( 0.089, 0.231) & 0.199 & ( 0.135, 0.263) \\
\beta_{\nm{weath}_4}   & 0.141 & ( 0.067, 0.215) & 0.133 & ( 0.066, 0.199) \\
\beta_{\nm{weath}_5}   & 0.176 & ( 0.103, 0.249) & 0.257 & ( 0.191, 0.323) \\[4pt]
\beta_{\nm{sleep}_2}   & 0.111 & ( 0.036, 0.187) & 0.135 & ( 0.068, 0.203) \\
\beta_{\nm{sleep}_3}   & 0.110 & ( 0.036, 0.184) & 0.176 & ( 0.110, 0.242) \\
\beta_{\nm{sleep}_4}   & 0.131 & ( 0.056, 0.205) & 0.162 & ( 0.095, 0.229) \\
\beta_{\nm{sleep}_5}   & 0.011 & (-0.064, 0.086) & 0.034 & (-0.033, 0.101) \\[4pt]
\beta_{\nm{act}_2}     & 0.135 & ( 0.060, 0.209) & 0.033 & (-0.033, 0.100) \\
\beta_{\nm{act}_3}     & 0.195 & ( 0.121, 0.269) & 0.203 & ( 0.137, 0.268) \\
\beta_{\nm{act}_4}     & 0.214 & ( 0.139, 0.290) & 0.254 & ( 0.186, 0.321) \\
\beta_{\nm{act}_5}     & 0.224 & ( 0.154, 0.294) & 0.158 & ( 0.095, 0.221) \\

\midrule[\heavyrulewidth]
\multicolumn{5}{l}{\footnotesize$^*$ statistically significant at 5\% level} \\
\bottomrule
\end{tabular}
\caption{Regression Coefficients of model 1 and model 2}\label{beta}
\end{table}
\end{document}