2024-10-02 08:18:07 发布
网友
在他的视频中,[Data analysis In Python with pandas](http://youtu.be/w26x-z-BdWQ?t=2h14s)中,Wes McKinney提出了一个系列方法,名为searchsorted(),它给出了一个值,并返回了序列穿过该值的索引。此功能似乎不再可用,是否有其他功能替代它?在
我相信这是由于Pandas 0.13.0中发生的重构,其中Pandas Series现在是NDFrame的子类,而不是ndarray参见this:
In [33]: import pandas as pd import numpy as np df = pd.DataFrame({'a':arange(10)}) df Out[33]: a 0 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9
[10行x 1列]
现在比较一下使用numpy数组会发生什么:
In [29]: temp = np.array(arange(10)) In [32]: temp.cumsum().searchsorted(11) Out[32]: 5
我相信这是由于Pandas 0.13.0中发生的重构,其中Pandas Series现在是NDFrame的子类,而不是ndarray参见this:
[10行x 1列]
^{pr2}$现在比较一下使用numpy数组会发生什么:
相关问题 更多 >
编程相关推荐