Pandas按行遍历DataFrame的3种方法

Pandas按行遍历DataFrame的3种方法

import pandas as pd
import numpy as np
import collections

df = pd.DataFrame(
    np.random.random(size=(100000, 4)), 
    columns=list('ABCD')
)
df.head(3)

.dataframe tbody tr th {
    vertical-align: top;
}

.dataframe thead th {
    text-align: right;
}

df.shape

(100000, 4)

1. df.iterrows()

使用方式

for idx, row in df.iterrows():
    print(idx, row)
    print(idx, row["A"], row["B"], row["C"], row["D"])
    break

0 A    0.897179
B    0.704997
C    0.419893
D    0.457423
Name: 0, dtype: float64
0 0.8971793669135368 0.7049968562422332 0.419892758556462 0.45742342531816327

时间耗费

%%time
result = collections.defaultdict(int)
for idx, row in df.iterrows():
    result[(row["A"], row["B"])] += row["A"] + row["B"]

Wall time: 8.43 s

2. df.itertuples()

使用方式

for row in df.itertuples():
    print(row)
    print(row.Index, row.A, row.B, row.C, row.D)
    break

Pandas(Index=0, A=0.8971793669135368, B=0.7049968562422332, C=0.419892758556462, D=0.45742342531816327)
0 0.8971793669135368 0.7049968562422332 0.419892758556462 0.45742342531816327

时间耗费

%%time
result = collections.defaultdict(int)
for row in df.itertuples():
    result[(row.A, row.B)] += row.A + row.B

Wall time: 327 ms

3. for+zip

使用方式

# 既不需要类型检查，也不需要构建namedtuple
# 缺点是需要挨个指定变量
for A, B in zip(df["A"], df["B"]):
    print(A, B)
    break

0.8971793669135368 0.7049968562422332

时间耗费

%%time
result = collections.defaultdict(int)
for A, B in zip(df["A"], df["B"]):
    result[(A, B)] += A + B

Wall time: 166 ms