Python中如何创建二维列表数组以处理CSV数据计算列均值

Hey there! Let's break this down step by step—first we'll fix that one-dimensional list problem, then cover how to compare your two CSVs and calculate column means.

1. Convert to a 2D List (Rows as Sublists) #

Your current code appends every single cell directly to my_list, which flattens everything into one big list. Instead, we need to create a sublist for each row, convert all cells in that row, then add the sublist to my_list.

Here's the revised code:

import csv

def conv(s):
    try:
        # Convert numerical strings (including scientific notation) to float
        return float(s)
    except ValueError:
        # Return non-numerical values (like dates/times) as-is
        return s

with open('Illia1.csv', newline='') as data:
    reader = csv.reader(data, delimiter=",")
    # Skip the first 8 header rows (cleaner than a while loop)
    for _ in range(8):
        next(reader, None)
    
    my_list = []
    for row in reader:
        # Convert each cell in the row and store as a sublist
        converted_row = [conv(cell) for cell in row]
        my_list.append(converted_row)

# Now my_list is 2D: each element is a row from the CSV
print(my_list)

This will give you output like:

[
    ['2001/01/24', '00:00.0', 1000.0, 1000.0, -0.0001, 0.0008],
    ['2001/01/25', '00:01.0', 2000.0, 2000.0, -0.0002, 0.0009],
    ...
]

2. Compare Two CSV Files for Differences #

Since your two CSVs have different headers but identical data content, we'll process both into 2D lists first, then compare row by row (and cell by cell if needed).

Here's a function to handle the comparison:

def compare_csvs(file1_path, file2_path, skip_rows=8):
    # Helper to load CSV into 2D list
    def load_csv(file_path):
        with open(file_path, newline='') as f:
            reader = csv.reader(f, delimiter=",")
            for _ in range(skip_rows):
                next(reader, None)
            return [[conv(cell) for cell in row] for row in reader]
    
    data1 = load_csv(file1_path)
    data2 = load_csv(file2_path)
    
    differences = []
    # Check row count first
    if len(data1) != len(data2):
        differences.append(f"Row count mismatch: {len(data1)} vs {len(data2)} rows")
    
    # Compare each row and cell
    for row_idx, (row1, row2) in enumerate(zip(data1, data2)):
        if len(row1) != len(row2):
            differences.append(f"Row {row_idx+1}: Column count mismatch ({len(row1)} vs {len(row2)})")
            continue
        for col_idx, (val1, val2) in enumerate(zip(row1, row2)):
            # For floats, use a small tolerance to avoid precision issues
            if isinstance(val1, float) and isinstance(val2, float):
                if abs(val1 - val2) > 1e-9:
                    differences.append(f"Row {row_idx+1}, Column {col_idx+1}: {val1} vs {val2}")
            elif val1 != val2:
                differences.append(f"Row {row_idx+1}, Column {col_idx+1}: {val1} vs {val2}")
    
    return differences

# Usage
diff_results = compare_csvs('Illia1.csv', 'Illia2.csv')
if diff_results:
    print("Found differences:")
    for diff in diff_results:
        print(f"- {diff}")
else:
    print("No differences between the two CSVs.")

3. Calculate Column Means #

Since the first two columns are dates/times, we'll skip those and calculate means for the numerical columns only:

def calculate_column_means(data):
    if not data:
        return []
    
    means = []
    # Start from column 2 (index 2) since first two are non-numerical
    for col_idx in range(2, len(data[0])):
        # Collect all float values in the column (skip any non-numeric entries)
        column_values = [row[col_idx] for row in data if isinstance(row[col_idx], float)]
        if column_values:
            mean = sum(column_values) / len(column_values)
            means.append(mean)
        else:
            means.append(None)  # Mark columns with no numeric data
    
    return means

# Usage
column_means = calculate_column_means(my_list)
print("Column means (starting from 3rd column):")
for idx, mean in enumerate(column_means, start=3):
    print(f"Column {idx}: {mean}")

内容的提问来源于stack exchange，提问作者IluxaKuk