# Why do I get unnecessary column space when using \multicolumn in type X columns?

Two of my columns have extra space before them which is not supposed to be there. I think the issue has to do with using \multicolumn inside a table with type X columns.

My source code:

\documentclass{IEEEtran}

\usepackage{fontspec}

\usepackage{tabularx}

\newcolumntype{C}{ >{ \arraybackslash \Centering } X }

\renewcommand \tabularxcolumn [1] {  >{ \centering } m{#1} }

\renewcommand{\tabularxcolumn}[1]{ >{\centering}m{#1} }

\usepackage{booktabs}

\usepackage{ragged2e}

\usepackage{makecell}

\usepackage{multirow}
\setlength\extrarowheight{10pt}

\usepackage{float}

\usepackage[defaultmathsizes, subdued, italic, LGRgreek, symbolre, symbolmisc]{mathastext}

\newcommand{\multcolhsize}[1]{\dimexpr #1\hsize + #1\tabcolsep + \tabcolsep \relax}

\begin{document}

\newcommand{\RERInteractionSumm}{\sum_{j \in \mathnormalbold{R}, \thinspace j \neq i} P_{R, \thinspace j}^{\adjustbox{raise=0.1\baselineskip}{*}} \thinspace w_{ij}}

\begin{table*}[htb]

% Table options

% Caption
\caption{Comparison of SDSCR obtained by}
\label{table:SDSCR_full_v_app_27}

% Center the table
\centering

\begin{tabularx}{\textwidth}{>{\hsize=0.9\hsize}C |>{\hsize=0.9\hsize}C |>{\hsize=0.9\hsize}C |>{\hsize=1.35\hsize}C |>{\hsize=1.35\hsize}C |>{\hsize=0.9\hsize}C |>{\hsize=0.9\hsize}C |>{\hsize=0.9\hsize}C |>{\hsize=0.9\hsize}C}

\toprule

$\left| V_{R, \thinspace i} \right|^{2}$ &  $\left| Z_{RR, \thinspace ii} \right|$& $P_{R, \thinspace i}^{\adjustbox{raise=0.05\baselineskip}{*}}$
& \multicolumn{2}{>{\hsize=\multcolhsize{2}}C|}{
$\displaystyle \RERInteractionSumm$
}
& \multicolumn{2}{>{\hsize=\multcolhsize{2}}C|}{
$\left| S_{eq, \thinspace i}^{\adjustbox{raise=0.05\baselineskip}{*}} \right|$
}
& \multicolumn{2}{>{\hsize=\multcolhsize{2}}C}{$SDSCR_{i}$}
\\
\Xhline{0.2mm}

$0.992$ & $0.0941$ & $3$ & $-1.05-0.095j$ & $-0.95-0.513j$ & $1.95$ & $2.11$ & $5.41$ & $4.99$
\\

$0.995$ & $0.0921$ & $3$ & $-1.08-0.108j$ & $-0.97-0.529j$ & $1.92$ & $2.10$ & $5.62$ & $5.16$
\\

$0.997$ & $0.0910$ & $3$ & $-1.09-0.115j$ & $-0.98-0.537j$ & $1.91$ & $2.09$ & $5.73$ & $5.25$
\\

$0.999$ & $0.0900$ & $3$ & $-1.11-0.124j$ & $-1.00-0.547j$ & $1.90$ & $2.08$ & $5.86$ & $5.35$
\\

$1.001$ & $0.0886$ & $3$ & $-1.12-0.132j$ & $-1.01-0.557j$ & $1.88$ & $2.07$ & $6.01$ & $5.46$
\\

$1.003$ & $0.0872$ & $3$ & $-1.14-0.143j$ & $-1.02-0.569j$ & $1.86$ & $2.06$ & $6.16$ & $5.58$
\\

$1.005$ & $0.0858$ & $3$ & $-1.16-0.154j$ & $-1.04-0.582j$ & $1.85$ & $2.05$ & $6.34$ & $5.72$
\\

$1.007$ & $0.0843$ & $3$ & $-1.18-0.166j$ & $-1.06-0.596j$ & $1.83$ & $2.04$ & $6.54$ & $5.87$
\\

$1.009$ & $0.0827$ & $3$ & $-1.21-0.180j$ & $-1.07-0.612j$ & $1.80$ & $2.02$ & $6.77$ & $6.04$
\\

$1.011$ & $0.0809$ & $3$ & $-1.23-0.196j$ & $-1.09-0.630j$ & $1.78$ & $2.01$ & $7.02$ & $6.23$
\\

\bottomrule

\end{tabularx}

\end{table*}

\end{document}

• The width calculations for the three header cells that span two columns each are incorrect. Anyway, it's not even necessary to calculate their widths. It suffices to the the basic c column type: \multicolumn{2}{c}{...} – Mico Nov 17 '18 at 21:11
• In addition to what @Mico said, I would never use tabularx for a data table of numbers, X columns are all about line breaking of text to a calculated width. – David Carlisle Nov 17 '18 at 21:41
• don't use math italic for multi-letter words, as you image shows they are spaced incorrectly for a word, use \mathrm{eq} and \mathrm{SDSCR} – David Carlisle Nov 17 '18 at 21:42

• Using a tabularx environment for the table at hand seems like vast overkill. Moreover, one of the hallmark capabilities of a tabularx environment, viz., automatic line breaking, is never used. I suggest you use an array environment or, if the tabular must absolutely, positively, span the width of both columns, a tabular* environment. With either environment, there's no need to calculate cell widths, etc.

• I would also like to suggest that you omit all vertical lines —-they’re not needed —- and be less generous with the value of \extrarowheight.

\documentclass{IEEEtran}

\usepackage{fontspec}
\usepackage{array,ragged2e,booktabs,float}
\newcolumntype{C}{>{\Centering\arraybackslash} X }
\setlength\extrarowheight{4pt}

\usepackage[defaultmathsizes, subdued, italic,
LGRgreek, symbolre, symbolmisc]{mathastext}
\newcommand{\RERInteractionSumm}{%
\sum_{j\in\mathbf{R},j\neq i} P_{R,j}^* w_{ij}}

\begin{document}
\begin{table*}[htb]
\caption{Comparison of SDSCR obtained by...}
\label{table:SDSCR_full_v_app_27}

\centering
%% First version: use an 'array' env.

$\begin{array}{@{} *{9}{c} @{}} \toprule | V_{R,i}|^{2} & | Z_{RR,ii}| & P_{R,i}^* & \multicolumn{2}{c}{\displaystyle\RERInteractionSumm}& \multicolumn{2}{c}{| S_{eq, i}^*|}& \multicolumn{2}{c@{}}{\mathrm{SDSCR}_{i}} \\ \cmidrule(r){1-1} \cmidrule(lr){2-2} \cmidrule(lr){3-3} \cmidrule(lr){4-5} \cmidrule(lr){6-7} \cmidrule(l){8-9} 0.992 & 0.0941 & 3 & -1.05-0.095j & -0.95-0.513j & 1.95 & 2.11 & 5.41 & 4.99 \\ 0.995 & 0.0921 & 3 & -1.08-0.108j & -0.97-0.529j & 1.92 & 2.10 & 5.62 & 5.16 \\ 0.997 & 0.0910 & 3 & -1.09-0.115j & -0.98-0.537j & 1.91 & 2.09 & 5.73 & 5.25 \\ 0.999 & 0.0900 & 3 & -1.11-0.124j & -1.00-0.547j & 1.90 & 2.08 & 5.86 & 5.35 \\ 1.001 & 0.0886 & 3 & -1.12-0.132j & -1.01-0.557j & 1.88 & 2.07 & 6.01 & 5.46 \\ 1.003 & 0.0872 & 3 & -1.14-0.143j & -1.02-0.569j & 1.86 & 2.06 & 6.16 & 5.58 \\ 1.005 & 0.0858 & 3 & -1.16-0.154j & -1.04-0.582j & 1.85 & 2.05 & 6.34 & 5.72 \\ 1.007 & 0.0843 & 3 & -1.18-0.166j & -1.06-0.596j & 1.83 & 2.04 & 6.54 & 5.87 \\ 1.009 & 0.0827 & 3 & -1.21-0.180j & -1.07-0.612j & 1.80 & 2.02 & 6.77 & 6.04 \\ 1.011 & 0.0809 & 3 & -1.23-0.196j & -1.09-0.630j & 1.78 & 2.01 & 7.02 & 6.23 \\ \bottomrule \end{array}$

\vspace{1cm}

%%% Second version: use a 'tabular*' env.
\setlength\tabcolsep{0pt}
\begin{tabular*}{\textwidth}{@{\extracolsep{\fill}} *{9}{c} }
\toprule
$| V_{R,i}|^{2}$ & $| Z_{RR,ii}|$ & $P_{R,i}^*$ &
\multicolumn{2}{c}{$\displaystyle\RERInteractionSumm$}&
\multicolumn{2}{c}{$| S_{eq, i}^*|$}&
\multicolumn{2}{c}{$\mathrm{SDSCR}_{i}$}
\\
\cmidrule{1-1}
\cmidrule{2-2}
\cmidrule{3-3}
\cmidrule{4-5}
\cmidrule{6-7}
\cmidrule{8-9}
$0.992$ & $0.0941$ & $3$ & $-1.05-0.095j$ & $-0.95-0.513j$ & $1.95$ & $2.11$ & $5.41$ & $4.99$
\\
$0.995$ & $0.0921$ & $3$ & $-1.08-0.108j$ & $-0.97-0.529j$ & $1.92$ & $2.10$ & $5.62$ & $5.16$
\\
$0.997$ & $0.0910$ & $3$ & $-1.09-0.115j$ & $-0.98-0.537j$ & $1.91$ & $2.09$ & $5.73$ & $5.25$
\\
$0.999$ & $0.0900$ & $3$ & $-1.11-0.124j$ & $-1.00-0.547j$ & $1.90$ & $2.08$ & $5.86$ & $5.35$
\\
$1.001$ & $0.0886$ & $3$ & $-1.12-0.132j$ & $-1.01-0.557j$ & $1.88$ & $2.07$ & $6.01$ & $5.46$
\\
$1.003$ & $0.0872$ & $3$ & $-1.14-0.143j$ & $-1.02-0.569j$ & $1.86$ & $2.06$ & $6.16$ & $5.58$
\\
$1.005$ & $0.0858$ & $3$ & $-1.16-0.154j$ & $-1.04-0.582j$ & $1.85$ & $2.05$ & $6.34$ & $5.72$
\\
$1.007$ & $0.0843$ & $3$ & $-1.18-0.166j$ & $-1.06-0.596j$ & $1.83$ & $2.04$ & $6.54$ & $5.87$
\\
$1.009$ & $0.0827$ & $3$ & $-1.21-0.180j$ & $-1.07-0.612j$ & $1.80$ & $2.02$ & $6.77$ & $6.04$
\\
$1.011$ & $0.0809$ & $3$ & $-1.23-0.196j$ & $-1.09-0.630j$ & $1.78$ & $2.01$ & $7.02$ & $6.23$
\\
\bottomrule
\end{tabular*}
\end{table*}

\end{document}

• +1 for not using my package:-) – David Carlisle Nov 17 '18 at 21:43