Question

假设我有一个数据框：

import pandas as pd
from random import randint

df = pd.DataFrame({'A': [randint(1, 9) for x in xrange(1000)],
                   'B': ...,
                   'C':....})

我要选择满足以下条件的行：如果至少X个连续的相邻行（在任一方向上）具有满足以下条件的A值，则选择行：abs（myRowAValue-meanAValueOfTheXNeighbors）

换句话说，我想在A值相当恒定的地方选择行。

我正在寻找最有效的“熊猫”方式。感谢您的帮助。

Answer 1

我不确定您的预期结果会是什么样，但是请问这是否可以帮到您（可能不是最高效的，也不是所有的熊猫）：

import pandas as pd
from random import randint
import numpy as np

df = pd.DataFrame({'A': [randint(1, 9) for x in range(1000)]})
neighbours = 10
tolerance = 2

nparray = np.array(df['A'])
nparray_len = len(nparray)

fbegin = [iterator for iterator, element in enumerate(nparray) if abs(element - np.average(nparray[:iterator+neighbours])) < tolerance and iterator < neighbours]
fmid = [iterator for iterator, element in enumerate(nparray) if abs(element - np.average(nparray[iterator-neighbours:iterator+neighbours])) < tolerance and iterator >= neighbours and iterator <= nparray_len - neighbours]
fend = [iterator for iterator, element in enumerate(nparray) if abs(element - np.average(nparray[iterator-neighbours:])) < tolerance and iterator > nparray_len - neighbours]
IDS = np.unique(np.array(fbegin + fmid + fend))
nparray[IDS]


df_constant = df.iloc[IDS]
print(df_constant)

熊猫：根据相邻行值的条件选择行

1 个答案: