如果你所有的文件都是相同的格式，而你只是想创建一个新的CSV，那么就不要使用pandas。

import csv
import pathlib

def write_to_writer(writer: csv.writer, path: pathlib.Path, skipheader=False) -> None:
    with path.open(newline="") as f:
        reader = csv.reader(f)
        if skipheader:
            next(reader)
        writer.writerows(reader)

folder_path = pathlib.Path('Some path here')
csv_files = [p for p in folder_path.iterdir() if p.suffix =='.csv']

with open('combined_data.csv', 'w', newline="") as fout:
    writer = csv.writer(fout)
    # handle first csv, with header
    first_path, *rest = csv_paths
    write_to_writer(writer, first_path, skipheader=False)
    for path in rest:
        write_to_writer(writer, path, skipheader=True)

这避免了循环中的pd.concat问题，即classic antipattern。
然而，如果你需要一个大的数据框（因为你要实际使用数据框），那么你的问题只需要添加到一个列表中，然后在最后进行协调即可解决：

import pathlib

import pandas as pd

folder_path = pathlib.Path('Some path here')
csv_paths = [p for p in folder_path.iterdir() if p.suffix =='.csv']

df = pd.concat([
    pd.read_csv(path, dtype={'stringbvariablename': str})
    for path in csv_paths
])

同样，这避免了反模式（因为.append在循环中执行到列表 * 是高效的 *）