1三、表格樣式建立

 

 

In [ ]:

'''
表格樣式建立

表格視覺樣式：Dataframe.style → 返回pandas.Styler對象的屬性，具備格式化和顯示Dataframe的有用方法

樣式建立：
① Styler.applymap：elementwise → 按元素方式處理Dataframe
② Styler.apply：column- / row- / table-wise → 按行/列處理Dataframe

'''

In [2]:

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
%matplotlib inline

In [3]:

# 樣式

df = pd.DataFrame(np.random.randn(10,4),columns=['a','b','c','d'])
sty = df.style
print(sty,type(sty))
# 查看樣式類型

sty
# 顯示樣式

 

<pandas.io.formats.style.Styler object at 0x000002B4D97A5E10> <class 'pandas.io.formats.style.Styler'>

Out[3]:

 

	a	b	c	d
0	0.994717	-1.76931	-1.04364	0.662782
1	0.507953	-0.751604	-3.20665	-0.00316762
2	-1.67345	0.777769	1.52595	-0.68124
3	1.39059	0.303007	1.36888	-1.40817
4	0.590499	1.25339	-1.843	-1.19178
5	0.907079	-0.752169	0.646874	-0.104434
6	-0.282473	-0.13618	0.425609	0.256449
7	0.317963	-0.188314	-0.395865	2.02621
8	-1.62159	1.22982	2.43284	-1.12136
9	0.73547	0.64652	-0.455337	0.923736

In [4]:

# 按元素處理樣式：style.applymap()

def color_neg_red(val):
    if val < 0:
        color = 'red'
    else:
        color = 'black'
    return('color:%s' % color)
df.style.applymap(color_neg_red)

# 建立樣式方法，使得小於0的數變成紅色
# style.applymap() → 自動調用其中的函數

Out[4]:

 

	a	b	c	d
0	0.994717	-1.76931	-1.04364	0.662782
1	0.507953	-0.751604	-3.20665	-0.00316762
2	-1.67345	0.777769	1.52595	-0.68124
3	1.39059	0.303007	1.36888	-1.40817
4	0.590499	1.25339	-1.843	-1.19178
5	0.907079	-0.752169	0.646874	-0.104434
6	-0.282473	-0.13618	0.425609	0.256449
7	0.317963	-0.188314	-0.395865	2.02621
8	-1.62159	1.22982	2.43284	-1.12136
9	0.73547	0.64652	-0.455337	0.923736

In [5]:

# 按行/列處理樣式：style.apply()

def highlight_max(s):
    is_max = s == s.max()
    #print(is_max) # 布爾型索引
    lst = []
    for v in is_max:
        if v:
            lst.append('background-color: yellow')
        else:
            lst.append('')
    return(lst)
df.style.apply(highlight_max, axis = 0, subset = ['b','c'])
# 建立樣式方法，每列最大值填充黃色
# axis：0爲列，1爲行，默認爲0
# subset：索引

Out[5]:

 

	a	b	c	d
0	0.994717	-1.76931	-1.04364	0.662782
1	0.507953	-0.751604	-3.20665	-0.00316762
2	-1.67345	0.777769	1.52595	-0.68124
3	1.39059	0.303007	1.36888	-1.40817
4	0.590499	1.25339	-1.843	-1.19178
5	0.907079	-0.752169	0.646874	-0.104434
6	-0.282473	-0.13618	0.425609	0.256449
7	0.317963	-0.188314	-0.395865	2.02621
8	-1.62159	1.22982	2.43284	-1.12136
9	0.73547	0.64652	-0.455337	0.923736

In [7]:

# 樣式索引、切片

df.style.apply(highlight_max, axis = 1, 
               subset = pd.IndexSlice[2:5,['b', 'd']])
# 經過pd.IndexSlice[]調用切片
# 也可：df[2:5].style.apply(highlight_max, subset = ['b', 'd']) → 先索引行再作樣式

Out[7]:

 

	a	b	c	d
0	0.994717	-1.76931	-1.04364	0.662782
1	0.507953	-0.751604	-3.20665	-0.00316762
2	-1.67345	0.777769	1.52595	-0.68124
3	1.39059	0.303007	1.36888	-1.40817
4	0.590499	1.25339	-1.843	-1.19178
5	0.907079	-0.752169	0.646874	-0.104434
6	-0.282473	-0.13618	0.425609	0.256449
7	0.317963	-0.188314	-0.395865	2.02621
8	-1.62159	1.22982	2.43284	-1.12136
9	0.73547	0.64652	-0.455337	0.923736

In [ ]: