基于多表列名实现MATLAB动态变量赋值的问题

问题描述 #

我需要基于数据集表格的列名创建动态变量，当前测试代码能生成目标变量名，但赋值为x^2，实际需求是将变量赋值为对应表格的列数据，期望生成类似imu1_accel_x0这样的变量，其值为对应表格列的全部数据。

原测试代码 #

varnammes = ["SpeedGoatT" "imu1_accel_x" "imu1_accel_y" "imu1_accel_z"  "imu1_gyro_x"   "imu1_gyro_y"   "imu1_gyro_z"   "imu1_orient_i" "imu1_orient_j" "imu1_orient_k" "imu1_orient_r" "imu2_accel_x"  "imu2_accel_y"  "imu2_accel_z"  "imu2_gyro_x"   "imu2_gyro_y"   "imu2_gyro_z"   "imu2_orient_i" "imu2_orient_j" "imu2_orient_k" "imu2_orient_r" "imu3_accel_x"  "imu3_accel_y"  "imu3_accel_z"  "imu3_gyro_x"   "imu3_gyro_y"   "imu3_gyro_z"   "imu3_orient_i" "imu3_orient_j" "imu3_orient_k" "imu3_orient_r" "imu4_accel_x"  "imu4_accel_y"  "imu4_accel_z"  "imu4_gyro_x"   "imu4_gyro_y"   "imu4_gyro_z"   "imu4_orient_i" "imu4_orient_j" "imu4_orient_k" "imu4_orient_r" "imu5_accel_x"  "imu5_accel_y"  "imu5_accel_z"  "imu5_gyro_x"   "imu5_gyro_y"   "imu5_gyro_z"   "imu5_orient_i" "imu5_orient_j" "imu5_orient_k" "imu5_orient_r" "imu6_accel_x"  "imu6_accel_y"  "imu6_accel_z"  "imu6_gyro_x"   "imu6_gyro_y"   "imu6_gyro_z"   "imu6_orient_i" "imu6_orient_j" "imu6_orient_k" "imu6_orient_r" "imu7_accel_x"  "imu7_accel_y"  "imu7_accel_z"  "imu7_gyro_x"   "imu7_gyro_y"   "imu7_gyro_z"   "imu7_orient_i" "imu7_orient_j" "imu7_orient_k" "imu7_orient_r"];
Tables = struct();
for i = 1:length(files)
    load(files(i))
    varName = sprintf('Table0%d', i-1); % Create a variable name
    Tables.(varName) = logsout.FileLogSignals{4}.Values;% Set the variable in the structure
    Tables.(varName) = timeseries2timetable(Tables.(varName));
    Tables.(varName) = timetable2table(Tables.(varName));
    Tables.(varName) = splitvars(Tables.(varName), 2);
    Tables.(varName).Properties.VariableNames(2:end) = varnammes;
    for x = 1:length(varnammes)
    % this is the problem:
    varName = [char(varnammes(x)) num2str(i-1)];
    eval([varName ' = x^2;']);
    %
    end
end

常规正确赋值写法 #

imu1_accel_x = Tables.Table00{:,3};
% 或者
imu1_accel_x = Tables.Table00.imu1_accel_x;

我尝试将eval语句改为以下代码但报错，需要正确的实现方式：

eval([varName ' = Tables.Table00.imu1_accel_x;']);

解决方案 #

1. 修复eval的实现方式 #

错误原因是硬编码了表格名Table00和列名，没有动态关联当前循环的表格和列。修改内层循环代码如下：

for x = 1:length(varnammes)
    % 生成带后缀的目标变量名
    targetVarName = [char(varnammes(x)) num2str(i-1)];
    % 动态获取当前循环的表格名
    currentTableName = sprintf('Table0%d', i-1);
    % 构造动态赋值语句
    eval([targetVarName ' = Tables.' currentTableName '.' char(varnammes(x)) ';']);
end

2. 更优方案：避免使用eval #

eval会降低代码可读性且调试困难，推荐用结构体统一管理动态变量：

% 提前创建结构体存储所有动态变量
DynamicVars = struct();

for i = 1:length(files)
    load(files(i))
    currentTableName = sprintf('Table0%d', i-1);
    Tables.(currentTableName) = logsout.FileLogSignals{4}.Values;
    Tables.(currentTableName) = timeseries2timetable(Tables.(currentTableName));
    Tables.(currentTableName) = timetable2table(Tables.(currentTableName));
    Tables.(currentTableName) = splitvars(Tables.(currentTableName), 2);
    Tables.(currentTableName).Properties.VariableNames(2:end) = varnammes;
    
    for x = 1:length(varnammes)
        varKey = [char(varnammes(x)) num2str(i-1)];
        % 将数据存入结构体，无需eval
        DynamicVars.(varKey) = Tables.(currentTableName).(char(varnammes(x)));
    end
end

后续可通过DynamicVars.imu1_accel_x0直接访问对应数据，比全局变量更安全易维护。

内容的提问来源于stack exchange，提问作者Tondre Benjamin D