Python Pandas 样式突出显示具有不同条件的每一列的特定单元格

Question

我正在尝试突出显示具有不同条件的每一列的特定单元格，它们的值与每一行的条件匹配。

下图是我想要实现的： The table I attempt to achieve

我搜索了 google 和 stackoverflow，但这些都不能满足我的要求。任何熟悉 Pandas Style 的人都可以提供帮助吗？

以下是我尝试过但失败的代码：

Ex1

import pandas as pd
df = pd.DataFrame([[10,3,1], [3,7,2], [2,4,4]], columns=list("ABC"))

def highlight(s):
    return ['background-color: yellow' if (v>2) else 'background-color: white' for v in s]
df.style.apply(highlight, axis=0)

Ex2

import pandas as pd
df = pd.DataFrame([[10,3,1], [3,7,2], [2,4,4]], columns=list("ABC"))

Column_limit = (df['A'] > 6) | (df['B'] > 2) | (df['C'] < 3)
df[Column_limit].style.applymap(lambda x: 'background-color: yellow', subset=pd.IndexSlice[:, ['A', 'C']])

Ex3

import pandas as pd
df = pd.DataFrame([[10,3,1], [3,7,2], [2,4,4]], columns=list("ABC"))

subsets = pd.IndexSlice[:, 'A']
df.style.applymap(lambda x: 'background-color: yellow', subset = subsets)

Answer 1

如果有相同数量的条件，比如一些列使用：

df = pd.DataFrame([[10,3,1], [3,7,2], [2,4,4]], columns=list("ABC"))

def highlight(x):
    c1 = 'background-color: yellow'

    # condition
    m = pd.concat([(x['A'] > 6), (x['B'] > 2), (x['C'] < 3)], axis=1)
    #print (m)
    #empty DataFrame of styles
    df1 = pd.DataFrame('', index=x.index, columns=x.columns)
    #set new columns by condition
    return df1.mask(m, c1)


df.style.apply(highlight, axis=None)

如果有很多列并且只需要处理其中的一部分：

def highlight(x):
    c1 = 'background-color: yellow'

    #empty DataFrame of styles
    df1 = pd.DataFrame('', index=x.index, columns=x.columns)
    #set new columns by condition
    
    df1.loc[(x['A'] > 6), 'A'] = c1
    df1.loc[(x['B'] > 2), 'B'] = c1
    df1.loc[(x['C'] < 3), 'C'] = c1
    
    return df1

df.style.apply(highlight, axis=None)

编辑：

如果需要指定所有掩码但在最后一步过滤器中仅使用某些列：

def highlight(x):
    c1 = 'background-color: yellow'

    #empty DataFrame of styles
    df1 = pd.DataFrame('', index=x.index, columns=x.columns)
    #set new columns by condition
    
    df1.loc[(x['A'] > 6), 'A'] = c1
    df1.loc[(x['B'] > 2), 'B'] = c1
    df1.loc[(x['C'] < 3), 'C'] = c1
    
    need = ['A','C']
    df1 = df1[need].reindex(x.columns, fill_value='', axis=1)
    return df1

或者去掉不需要的掩码：

def highlight(x):
    c1 = 'background-color: yellow'

    #empty DataFrame of styles
    df1 = pd.DataFrame('', index=x.index, columns=x.columns)
    #set new columns by condition
    
    df1.loc[(x['A'] > 6), 'A'] = c1
    df1.loc[(x['C'] < 3), 'C'] = c1
    
    return df1

df.style.apply(highlight, axis=None)