Matlab实现按第一列数值确定行位置向表格添加行并处理覆盖逻辑

Here's how you can modify your MATLAB code to achieve the desired result—aligning rows by the first column (file number), filling missing IDs with NaN (or 0), and overwriting existing rows with new data:

% Step 1: Load existing CSV file and preserve variable names
file = 'directory/test.csv';
alpha_old = readtable(file);
% Define the ID column (first column in your tables)
id_col = alpha_old.Properties.VariableNames{1}; % Should be 'alpha21' in your case

% Step 2: Prepare new data table (your existing code here)
beta = array2table(newdata);
beta.Properties.VariableNames = {'alpha21' 'alpha22' 'alpha23' 'alpha24' };

% Step 3: Create a complete range of IDs from min to max of both tables
all_ids = unique([alpha_old.(id_col); beta.(id_col)]);
min_id = min(all_ids);
max_id = max(all_ids);
full_ids = (min_id:max_id)'; % Column vector of all IDs from min to max

% Step 4: Create base table with full IDs and NaN-filled columns
base_table = array2table(full_ids, 'VariableNames', {id_col});
% Add other columns from the original table, initialized to NaN
other_columns = setdiff(alpha_old.Properties.VariableNames, id_col);
for col = other_columns
    base_table.(col{1}) = NaN(size(full_ids));
end

% Step 5: Populate base table with old data
% Find indices where old IDs match full IDs
idx_old = ismember(base_table.(id_col), alpha_old.(id_col));
% Fill in old data for matching IDs
base_table(idx_old, other_columns) = alpha_old(:, other_columns);

% Step 6: Overwrite with new data (if IDs exist in new table)
% Find indices where new IDs match full IDs
idx_new = ismember(base_table.(id_col), beta.(id_col));
% Get rows from beta that match full IDs, in the correct order
beta_matched = beta(ismember(beta.(id_col), base_table.(id_col)), :);
% Overwrite base table rows with new data
base_table(idx_new, :) = beta_matched;

% Step 7: Replace NaN with 0 if desired (uncomment below)
% base_table(isnan(base_table)) = 0;

% Step 8: Save the merged table back to CSV
writetable(base_table, file, 'WriteVariableNames', 0);

Key Details: #

• Preserving Variable Names: We skip converting the table to an array and back (like your original code did) to keep column names intact.
• Complete ID Range: We generate every integer ID from the smallest to largest present in both tables, ensuring no gaps in numbering.
• Overwrite Logic: New data takes priority—if an ID exists in both tables, the new row replaces the old one entirely.
• Missing Values: Rows for IDs not present in either table are filled with NaN; uncomment the final line in step 7 to replace NaN with 0 if needed.

Example Output: #

Using your sample data, the resulting table will look like this:

alpha21 alpha22 alpha23 alpha24
_______ __________ __________ __________
1       2.2204e-14 2.2204e-14 2.2204e-14
2       2.2204e-14 2.2204e-14 2.2204e-14
3       NaN        NaN        NaN
4       2.2204e-14 2.2204e-14 2.2204e-14
5       2.2204e-14 2.2204e-14 2.2204e-14

内容的提问来源于stack exchange，提问作者Mosawi