首先看一下map函数的运行规则
date = {'age': 15, 'name': 'lili', 'sex': 'girl'}
delete_module = ['name', 'sex']
a = map(lambda x: date.pop(x), delete_module)
print(date)
p = list(a)
print(date)
运行结果: 只有对map的结果a做了调用(list)才会对date做处理
{‘age’: 15, ‘name’: ‘lili’, ‘sex’: ‘girl’}
{‘age’: 15}
由此当对map结果多步处理时,处理完第一步时由于内存释放所造成进一步的问题
例子
import pandas as pd
d = {'p': {'a': [1, 2, 3, 4], 'b': [4, 5, 6, 7]}}
p = map(lambda x: d['p'], d)
print(list(p))
a = map(lambda r: r['a'], p)
print(list(a))
b = map(lambda x: x['b'], p)
list(b)
运行结果 当list§后,p所保留的数据也就释放了
[{‘a’: [1, 2, 3, 4], ‘b’: [4, 5, 6, 7]}]
[]
[]
import pandas as pd
import numpy as np
d = {'p': {'a': [1, 2, 3, 4], 'b': [4, 5, 6, 7]}}
p = map(lambda x: d['p'], d)
a = map(lambda r: r['a'], p)
print(list(a))
b = map(lambda x: x['b'], p)
list(b)
[[1, 2, 3, 4]]
[]
解决方法:
目前只想到了深拷贝的方法
import pandas as pd
import numpy as np
import copy
d = {'p': {'a': [1, 2, 3, 4], 'b': [4, 5, 6, 7]}}
p = map(lambda x: d['p'], d)
x=copy.deepcopy(p)
a = map(lambda r: r['a'], p)
print(list(a))
b = map(lambda y: y['b'], x)
list(b)