我似乎无法修复这个小问题，即单词“MS-DNS”下方的行与单词左侧对齐，而不是与单词右侧对齐。这是 LaTeX 代码
\documentclass[11pt,a4paper]{article}
\usepackage{amssymb, amsmath, bm}
\usepackage[flushleft]{threeparttable}
\usepackage{float, booktabs, makecell, caption, tabularx}
\usepackage{siunitx}
\begin{document}
\begin{table}[htb!]
\caption{\textbf{Filtered errors MS-DNS model and extensions with macro
factors}}
\label{table:filtered_errors_macro}
\centering
\begin{threeparttable}
\renewcommand{\TPTminimum}{\linewidth}
\makebox[\linewidth]{%
\begin{tabular}{@{} l *{8}{S[table-format=1.3]} @{}}
\toprule
\hline
& \multicolumn{2}{c}{MS-DNS}
& \multicolumn{2}{c}{MS-DRA}
& \multicolumn{2}{c}{MS-TVTP}
& \multicolumn{2}{c@{}}{MS-DNS-X} \\
\cmidrule(l){4-5} \cmidrule(l){6-7} \cmidrule(l){8-9} \cmidrule(l){10-11}
Maturity & {Mean} & {Std.\ dev.}
& {Mean} & {Std.\ dev.}
& {Mean} & {Std.\ dev.}
& {Mean} & {Std.\ dev.} \\
\midrule
3 & -0.066 & 0.063 & -0.070 & 0.345 & -0.139 & 0.330 & -0.229 & 1.608 \\
6 & 0.027 & 0.047 & 0.003 & 0.325 & -0.049 & 0.257 & -0.327 & 3.440 \\
12 & 0.076 & 0.072 & 0.037 & 0.334 & 0.014 & 0.226 & -0.556 & 6.140 \\
24 & 0.025 & 0.025 & -0.002 & 0.330 & 0.020 & 0.228 & -0.898 & 8.853 \\
36 & -0.034 & 0.037 & -0.038 & 0.324 & 0.012 & 0.227 & -1.059 & 9.544 \\
60 & -0.074 & 0.055 & -0.068 & 0.314 & 0.000 & 0.225 & -1.107 & 8.643 \\
84 & -0.032 & 0.026 & -0.050 & 0.304 & 0.010 & 0.222 & -1.013 & 7.118 \\
120 & 0.079 & 0.057 & 0.009 & 0.305 & 0.044 & 0.229 & -0.834 & 5.276\\
\addlinespace
Mean & 0.000 & 0.048 & -0.022 & 0.323 & -0.011 & 0.243 & -0.753 & 6.328 \\
Median & -0.004 & 0.051 & -0.020 & 0.324 & 0.011 & 0.228 & -0.866 & 6.629 \\
Lower & & & 5 & 0 & 5 & 0 & 0 & 0
\\
\bottomrule
\end{tabular}
}
\smallskip
\begin{tablenotes}[flushleft]\footnotesize\smallskip
\item Note: This table reports filtered errors from the MS-DNS model and its
extensions with macro factors. The filtered errors are defined as the residuals
of the measurement equation and are formed by taking the difference of the
observed yield and its filtered estimate obtained by the Kim filter. For each
maturity, I present the mean and standard deviation (\textit{Std. dev.}).
Moreover, I show the mean, median and number of maturities for which the mean
of the extended macro model is lower than the mean of the MS-DNS model in
absolute terms (\textit{Lower}).
\end{tablenotes}
\end{threeparttable}
\end{table}
\end{document}
LaTeX 代码的输出希望能清楚地表明问题是什么：
先感谢您。
答案1
改变线路
\cmidrule(l){4-5} \cmidrule(l){6-7} \cmidrule(l){8-9} \cmidrule(l){10-11}
到
\cmidrule(l){2-3} \cmidrule(l){4-5} \cmidrule(l){6-7} \cmidrule(l){8-9}
括号中的数字对应于每条规则应跨越的列。您的列从 1 到 9，因此\cmidrule(l){10-11}会出现有关插入额外对齐制表符的错误，这是 TeX 表示您的数字“超出范围”的方式。
新的输出符合预期：
编辑：正如 leandriis 在评论中指出的那样，你还应该替换
\begin{tabular}{@{} l *{8}{S[table-format=1.3]} @{}}
由其中之一
\begin{tabular}{@{} l *{8}{S[table-format=-1.3]} @{}}
\begin{tabular}{@{} l *{4}{S[table-format=-1.3]S[table-format=1.3]} @{}}
为减号保留额外的空间（在所有或仅在“平均值”列中），从而消除警告Overfull \hbox。
答案2
您想要\cmidrule{2-3}等等：通过该输入，您会收到一条关于“列太多”的错误消息；TeX 会尽力恢复，但应该会修复。
我不会让表格超出边距，而是将其排版\small：
\documentclass[11pt,a4paper]{article}
\usepackage{amssymb, amsmath, bm}
\usepackage[flushleft]{threeparttable}
\usepackage{float, booktabs, makecell, caption, tabularx}
\usepackage{siunitx}
\begin{document}
\begin{table}[htb!]
\caption{Filtered errors MS-DNS model and extensions with macro factors}
\label{table:filtered_errors_macro}
\begin{threeparttable}
\centering
\small
\setlength\tabcolsep{0pt}
\begin{tabular*}{\textwidth}{
@{}
l
@{\extracolsep{\fill}}
*{4}{S[table-format=-1.3]S[table-format=1.3]} @{}}
\toprule
Maturity
& \multicolumn{2}{c}{MS-DNS}
& \multicolumn{2}{c}{MS-DRA}
& \multicolumn{2}{c}{MS-TVTP}
& \multicolumn{2}{c@{}}{MS-DNS-X} \\
\cmidrule{2-3} \cmidrule{4-5} \cmidrule{6-7} \cmidrule{8-9}
& {Mean} & {Std.\ dev.}
& {Mean} & {Std.\ dev.}
& {Mean} & {Std.\ dev.}
& {Mean} & {Std.\ dev.} \\
\midrule
3 & -0.066 & 0.063 & -0.070 & 0.345 & -0.139 & 0.330 & -0.229 & 1.608 \\
6 & 0.027 & 0.047 & 0.003 & 0.325 & -0.049 & 0.257 & -0.327 & 3.440 \\
12 & 0.076 & 0.072 & 0.037 & 0.334 & 0.014 & 0.226 & -0.556 & 6.140 \\
24 & 0.025 & 0.025 & -0.002 & 0.330 & 0.020 & 0.228 & -0.898 & 8.853 \\
36 & -0.034 & 0.037 & -0.038 & 0.324 & 0.012 & 0.227 & -1.059 & 9.544 \\
60 & -0.074 & 0.055 & -0.068 & 0.314 & 0.000 & 0.225 & -1.107 & 8.643 \\
84 & -0.032 & 0.026 & -0.050 & 0.304 & 0.010 & 0.222 & -1.013 & 7.118 \\
120 & 0.079 & 0.057 & 0.009 & 0.305 & 0.044 & 0.229 & -0.834 & 5.276\\
\addlinespace
Mean & 0.000 & 0.048 & -0.022 & 0.323 & -0.011 & 0.243 & -0.753 & 6.328 \\
Median & -0.004 & 0.051 & -0.020 & 0.324 & 0.011 & 0.228 & -0.866 & 6.629 \\
Lower & & & 5 & 0 & 5 & 0 & 0 & 0
\\
\bottomrule
\end{tabular*}
\begin{tablenotes}[flushleft]\footnotesize\smallskip
\item \hspace*{-\fontdimen2\font}Note: This table reports filtered errors from the MS-DNS model and its
extensions with macro factors. The filtered errors are defined as the residuals
of the measurement equation and are formed by taking the difference of the
observed yield and its filtered estimate obtained by the Kim filter. For each
maturity, I present the mean and standard deviation (\textit{Std.\ dev.}).
Moreover, I show the mean, median and number of maturities for which the mean
of the extended macro model is lower than the mean of the MS-DNS model in
absolute terms (\textit{Lower}).
\end{tablenotes}
\end{threeparttable}
\end{table}
\end{document}
您可以\small通过减少列标题来避免：
\documentclass[11pt,a4paper]{article}
\usepackage{amssymb, amsmath, bm}
\usepackage[flushleft]{threeparttable}
\usepackage{float, booktabs, makecell, caption, tabularx}
\usepackage{siunitx}
\begin{document}
\begin{table}[htb!]
\caption{Filtered errors MS-DNS model and extensions with macro factors}
\label{table:filtered_errors_macro}
\begin{threeparttable}
\centering
\setlength\tabcolsep{0pt}
\begin{tabular*}{\textwidth}{
@{}
l
@{\extracolsep{\fill}}
*{4}{S[table-format=-1.3]S[table-format=1.3]} @{}}
\toprule
Maturity
& \multicolumn{2}{c}{MS-DNS}
& \multicolumn{2}{c}{MS-DRA}
& \multicolumn{2}{c}{MS-TVTP}
& \multicolumn{2}{c@{}}{MS-DNS-X} \\
\cmidrule{2-3} \cmidrule{4-5} \cmidrule{6-7} \cmidrule{8-9}
& {M} & {SD}
& {M} & {SD}
& {M} & {SD}
& {M} & {SD} \\
\midrule
3 & -0.066 & 0.063 & -0.070 & 0.345 & -0.139 & 0.330 & -0.229 & 1.608 \\
6 & 0.027 & 0.047 & 0.003 & 0.325 & -0.049 & 0.257 & -0.327 & 3.440 \\
12 & 0.076 & 0.072 & 0.037 & 0.334 & 0.014 & 0.226 & -0.556 & 6.140 \\
24 & 0.025 & 0.025 & -0.002 & 0.330 & 0.020 & 0.228 & -0.898 & 8.853 \\
36 & -0.034 & 0.037 & -0.038 & 0.324 & 0.012 & 0.227 & -1.059 & 9.544 \\
60 & -0.074 & 0.055 & -0.068 & 0.314 & 0.000 & 0.225 & -1.107 & 8.643 \\
84 & -0.032 & 0.026 & -0.050 & 0.304 & 0.010 & 0.222 & -1.013 & 7.118 \\
120 & 0.079 & 0.057 & 0.009 & 0.305 & 0.044 & 0.229 & -0.834 & 5.276\\
\addlinespace
Mean & 0.000 & 0.048 & -0.022 & 0.323 & -0.011 & 0.243 & -0.753 & 6.328 \\
Median & -0.004 & 0.051 & -0.020 & 0.324 & 0.011 & 0.228 & -0.866 & 6.629 \\
Lower & & & 5 & 0 & 5 & 0 & 0 & 0
\\
\bottomrule
\end{tabular*}
\begin{tablenotes}[flushleft]\footnotesize\smallskip
\item \hspace*{-\fontdimen2\font}Note: This table reports filtered errors from
the MS-DNS model and its extensions with macro factors. The filtered errors are
defined as the residuals of the measurement equation and are formed by taking
the difference of the observed yield and its filtered estimate obtained by the
Kim filter. For each maturity, I present the mean (M,~\textit{Mean}) and standard
deviation (SD,~\textit{Std.\ dev.}). Moreover, I show the mean, median and number
of maturities for which the mean of the extended macro model is lower than the
mean of the MS-DNS model in absolute terms (\textit{Lower}).
\end{tablenotes}
\end{threeparttable}
\end{table}
\end{document}
