float 32 数据和 float 64 数据的内存使用相同吗？

Question

我尝试使用 tracemalloc 来跟踪我的编码的内存使用情况，我发现对于 float 32 数据和 float 64 数据，它们具有相同的内存使用峰值。为什么？

输出截图：

float64:

import tracemalloc
import numpy as np
tracemalloc.start()
arr = np.zeros((100000000,), dtype=np.float64)
for i in range(100000000):
    arr[i] = i
current, peak = tracemalloc.get_traced_memory()
#current memory is the memory the code is currently using and 
#peak memory is the maximum space the program used while executing.
tracemalloc.stop()
print(f"Current memory usage is {current / 10**6}MB; Peak was {peak / 10**6}MB")

输出：

Current memory usage is 800.14981MB; Peak was 800.160145MB

float32:

import tracemalloc
import numpy as np
tracemalloc.start()
arr = np.zeros((100000000,), dtype=np.float32)
arr = arr.astype(np.float32)
for i in range(100000000):
    arr[i] = i
current, peak = tracemalloc.get_traced_memory()
tracemalloc.stop()
print(f"Current memory usage is {current / 10**6}MB; Peak was {peak / 10**6}MB")

输出：

Current memory usage is 400.149626MB; Peak was 800.149468MB

Answer 1

您可以看到这 2 个片段并不相同。第二个片段中有这个附加声明

arr = arr.astype(np.float32)

该语句（即使 arr 已经有类型 np.float32）：

复制数组（占用与旧数组相同的内存量）
然后将新数组赋值给arr,
然后丢弃旧数组。

因此内存峰值翻倍。

float 32 数据和 float 64 数据的内存使用相同吗？

Same Memory Usage for float 32 data and float 64 data?

python

memory

out-of-memory

python-3.x