使用 pandas to_csv 時刪除標題中的前導逗號

Question

默認情況下 to_csv 寫一個 CSV 像

,a,b,c
0,0.0,0.0,0.0
1,0.0,0.0,0.0
2,0.0,0.0,0.0

但我希望它這樣寫：

a,b,c
0,0.0,0.0,0.0
1,0.0,0.0,0.0
2,0.0,0.0,0.0

我如何實現這一目標？ 我不能設置index=False因為我想保留索引。 我只想刪除前導逗號。

df = pd.DataFrame(np.zeros((3,3)), columns = ['a','b','c'])
df.to_csv("test.csv") # this results in the first example above.

Answer 1

可以先只寫沒有索引的列，然后在追加模式下寫沒有標題的數據：

df = pd.DataFrame(np.zeros((3,3)), columns = ['a','b','c'], index=list('XYZ'))

pd.DataFrame(columns=df.columns).to_csv("test.csv", index=False)
#alternative for empty df
#df.iloc[:0].to_csv("test.csv", index=False)
df.to_csv("test.csv", header=None, mode='a')

df = pd.read_csv("test.csv")
print (df)
     a    b    c
X  0.0  0.0  0.0
Y  0.0  0.0  0.0
Z  0.0  0.0  0.0

Answer 2

只需為您的索引設置一個名稱： df.index.name = 'blah' 。 此名稱將作為名字出現在標題中。

import numpy as np
import pandas as pd

df = pd.DataFrame(np.zeros((3,3)), columns = ['a','b','c'])
df.index.name = 'my_index'
print(df.to_csv())

產量

my_index,a,b,c
0,0.0,0.0,0.0
1,0.0,0.0,0.0
2,0.0,0.0,0.0

但是，如果（根據您的評論）您希望在標題中有 3 個以逗號分隔的名稱，而在 csv 的行中有 4 個以逗號分隔的值，則您必須手工制作它。 但是它不符合任何 csv 標准格式。

Answer 3

或者，嘗試重置索引，使其成為數據框中的一列，名為index 。 這也適用於多個索引。

df = df.reset_index()
df.to_csv('output.csv', index = False)