map和applymap及apply的区别

map和applymap及apply的区别

1. 数据

import pandas as pd
import numpy as np

frame = pd.DataFrame(np.random.rand(4, 3), columns = list('abc'), index = ['Utah', 'Ohio', 'Texas', 'Oregon'])
print(frame)
# 输出如下:
#                a         b         c
# Utah    0.443188  0.919623  0.550259
# Ohio    0.013923  0.557696  0.723975
# Texas   0.865469  0.720604  0.081306
# Oregon  0.506174  0.212421  0.061561

2. apply()

apply()是pandas里DataFrame的函数，可以针对DataFrame中的行数据或列数据应用操作。

注意：这里的apply是应用到每一行或者每一列操作，有专门的axis可以指定，默认是axis=0.

func = lambda x: x.max() - x.min()
print(frame.apply(func))
# 输出如下:
# a    0.851545
# b    0.707202
# c    0.662415
# dtype: float64

# --------------------------
df_im_data.loc[:,["from_user", "to_user"]].head()

#           from_user   to_user
#   id		
# 1094736	sso-1361713	tuliu001
# 1094737	tuliu001	sso-1361713
# 1094738	sso-1367729	sc-1131
# 1094739	sso-1250193	tuliu002
# 1094741	sso-1367729	tuliu002

def get_two_user_code(user1, user2):
    s = [user1, user2]
    l = sorted(s)
    code = "@".join(l)
    return code
    
c = df_im_data.loc[:,["from_user", "to_user"]].apply(lambda x: get_two_user_code(x[0], x[1]), axis=1)

# id
# 1094736    sso-1361713@tuliu001
# 1094737    sso-1361713@tuliu001
# 1094738     sc-1131@sso-1367729
# 1094739    sso-1250193@tuliu002
# 1094741    sso-1367729@tuliu002
# dtype: object

3. applymap()

apply()是pandas里DataFrame的函数，但是它对DataFrame中所有的元素应用操作。

func = lambda x: f'{x:.2f}%'
print(frame.applymap(func))

# 输出如下:
#             a      b      c
# Utah    0.34%  0.43%  0.67%
# Ohio    0.75%  0.50%  0.14%
# Texas   0.68%  0.28%  0.90%
# Oregon  0.05%  0.86%  0.78%

4. map()

只能直接对序列元素的操作，所以必须对DataFrame的某列 (只针对于列) 应用操作。

func = lambda x: f'{x:.2f}%'
print(frame['a'].map(func))
# 输出如下:
# Utah      0.65%
# Ohio      0.90%
# Texas     0.09%
# Oregon    0.72%
# Name: a, dtype: object