我如何对一个panda.DataFrame或panda.Seriesin place应用矢量化操作？我只找到了创建和返回副本的方法。
我主要关心的是减少资源的使用，但如果能知道它是否可能，即使它不是特别有效，那也是很好的。
按照目前的情况，您可以执行以下操作：

import pandas as pd

df = pd.DataFrame({'a': [1,2,3], 'b': [4,5,6]})
#  df
#    a  b
# 0  1  4
# 1  2  5
# 2  3  6

df['c'] = df['a'] + df['b']
#  df
#    a  b  c
# 0  1  4  5
# 1  2  5  7
# 2  3  6  9

# Or like this
df['c'] = np.log(df['a'])
#  df
#    a  b         c
# 0  1  4  0.000000
# 1  2  5  0.693147
# 2  3  6  1.098612

# Or these methods:
# df['c'] = df['a'].apply(np.log)
# df['c'] = np.vectorize(np.log)(df['a'])

但是，我想做一些类似于R中的data.table所能做的事情。

df[c = np.log(a)]
#  df
#    a  b         c
# 0  1  4  0.000000
# 1  2  5  0.693147
# 2  3  6  1.098612

# or even
df['a'].apply(np.log, inplace=True) # which doesn't exist
# so that column 'a' were transformed in place
# df[['a','b']]
#           a  b
# 0  0.000000  4
# 1  0.693147  5
# 2  1.098612  6