Question

下面是我正在处理的数据片段，有数千行和其他列。我必须更改＆＃39;列Y＆＃39;根据以下条件在第X栏＆＃39;。

如果X列是＆＃34; FIRST＆＃34; ：

细胞＃1 =上皮细胞

细胞＃2 =神经

如果X列是＆＃34; SECOND＆＃34;：

细胞＃1 =内皮细胞

细胞＃2 =肌肉

数据帧：

 Column X            Column Y        
    FIRST               cell#1
    FIRST               A
    FIRST               cell#2
    FIRST               C
    SECOND              N
    SECOND              V
    SECOND              cell#1
    SECOND              cell#2

代码：

for row in df['Column X']:
    if row == "FIRST":
        df.loc[(df['Column Y']== "cell#1"), 'Column Y'] = "epithelial"
        df.loc[(df['Column Y']== "cell#2"), 'Column Y'] = "nerve"
    elif row == "SECOND":
        df.loc[(df['Column Y']== "cell#1"), 'Column Y'] = "endothelial"
        df.loc[(df['Column Y']== "cell#2"), 'Column Y'] = "muscle"
    else:
        pass

上面的代码不起作用，行的条件==＆＃39; FIRST＆＃39;适用于整个数据框，并忽略行==＆＃39; SECOND＆＃34;的条件。请帮忙。

预期结果：

Column X            Column Y        
    FIRST               epithelial
    FIRST               A
    FIRST               nerve
    FIRST               C
    SECOND              N
    SECOND              V
    SECOND              endothelial
    SECOND              muscle

我上面代码的输出（不正确）：

Column X            Column Y        
    FIRST               epithelial
    FIRST               A
    FIRST               nerve
    FIRST               C
    SECOND              N
    SECOND              V
    SECOND              epithelial
    SECOND              nerve

Y列中的最后两行应该是＆＃34;内皮细胞＆＃34;和＆＃34;肌肉＆＃34;，不＆＃34;上皮＆＃34;和＃34;神经＆＃34;

Answer 1

这是一种方式。请注意，不需要循环。为了方便和性能，许多pandas操作都被矢量化。

import pandas as pd

df = pd.DataFrame([['FIRST', 'cell#1'], ['FIRST', 'A'],
                   ['FIRST', 'cell#2'], ['FIRST', 'C'],
                   ['SECOND', 'N'], ['SECOND', 'V'],
                   ['SECOND', 'cell#1'], ['SECOND', 'cell#2']],
                  columns=['X', 'Y'])

df.loc[(df.X == 'FIRST') & (df.Y == 'cell#1'), 'Y'] = 'epithelial'
df.loc[(df.X == 'FIRST') & (df.Y == 'cell#2'), 'Y'] = 'nerve'
df.loc[(df.X == 'SECOND') & (df.Y == 'cell#1'), 'Y'] = 'endothelial'
df.loc[(df.X == 'SECOND') & (df.Y == 'cell#2'), 'Y'] = 'muscle'

#         X            Y
# 0   FIRST   epithelial
# 1   FIRST            A
# 2   FIRST        nerve
# 3   FIRST            C
# 4  SECOND            N
# 5  SECOND            V
# 6  SECOND  endothelial
# 7  SECOND       muscle

Answer 2

我自己去了解grouping in Pandas我自己的更多信息，并感到惊讶的是我找不到一种优雅的方式来做这件事。

我想出的是：

import pandas as pd

df = pd.DataFrame({'Column X': ['FIRST', 'FIRST', 'FIRST', 'FIRST', 'SECOND', 'SECOND', 'SECOND', 'SECOND'], 'Column Y': ['cell#1', 'A', 'cell#2', 'C', 'N', 'V', 'cell#1', 'cell#2']})

def f(group):
    y = group['Column Y']
    key = group['Column X'].iloc[0]
    if key == 'FIRST':
        y[y == 'cell#1'] = 'epithelial'
        y[y == 'cell#2'] = 'nerve'
    elif key == 'SECOND':
        y[y == 'cell#1'] = 'endothelial'
        y[y == 'cell#2'] = 'muscle'
    return group

df.groupby('Column X').apply(f)

但是这需要从分组列再次获取密钥，将其传递给f会更简单。

应用if语句来替换Y列中的数据＆＃39;对应于＆＃39;列X＆＃39;中的不同行

2 个答案: