Question

有没有办法在Python中找到两个时间戳数组的匹配元素？

我正在尝试为两个数组执行此操作：

import numpy as np
import pandas

xrp[:,0]
Out[135]: 
array([Timestamp('2018-03-08 00:00:00'), Timestamp('2018-03-07 00:00:00'),
       Timestamp('2018-03-06 00:00:00'), ...,
       Timestamp('2013-08-06 00:00:00'), Timestamp('2013-08-05 00:00:00'),
       Timestamp('2013-08-04 00:00:00')], dtype=object)

btc[:,0]
Out[136]: 
array([Timestamp('2018-03-08 00:00:00'), Timestamp('2018-03-07 00:00:00'),
       Timestamp('2018-03-06 00:00:00'), ...,
       Timestamp('2013-04-30 00:00:00'), Timestamp('2013-04-29 00:00:00'),
       Timestamp('2013-04-28 00:00:00')], dtype=object)

val = np.where(btc[:,0]==xrp[:,0])
__main__:1: DeprecationWarning: elementwise == comparison failed; this will raise an error in the future.

val
Out[138]: (array([], dtype=int64),)

似乎np.where和Timestamps不是很兼容，或者过去常常不再存在。

如果可能，我想在没有循环的情况下这样做。

每个数组的形状为：

xrp[:,0].shape
Out[139]: (1678,)

btc[:,0].shape
Out[140]: (1776,)

Answer 1

Let a and b be your arrays

i, j = np.where(a == b[:, np.newaxis])

print(a[i])
print(b[j])

pandas时间戳与np.where比较

1 个答案: