Pandas | 手册

%%html
<style>
@import url('https://fonts.googleapis.com/css?family=Ewert|Roboto&effect=3d|ice|');
body {background-color: gainsboro;} 
a {color: #37c9e1; font-family: 'Roboto';} 
h1 {color: #37c9e1; font-family: 'Orbitron'; text-shadow: 4px 4px 4px #aaa;} 
h2, h3 {color: slategray; font-family: 'Orbitron'; text-shadow: 4px 4px 4px #aaa;}
h4 {color: #818286; font-family: 'Roboto';}
span {font-family:'Roboto'; color:black; text-shadow: 5px 5px 5px #aaa;}  
div.output_area pre{font-family:'Roboto'; font-size:110%; color:lightblue;}      
</style>

🌑 Setup and Read Dataset

import pandas as pd

dataPath = '../input/zoo-animal-classification/'
animalClass = pd.read_csv(dataPath+'class.csv')
zoo = pd.read_csv(dataPath+'zoo.csv')

🌒 Data Structure

The type of rows and columns of ‘DataFrame’ is “Series”

# 1D(base) --> Series  
row = zoo.iloc[0]
col = zoo.eggs
print(type(row))
print(type(col))
# 2D(base) --> DataFram
print(type(zoo))

<class 'pandas.core.series.Series'>
<class 'pandas.core.series.Series'>
<class 'pandas.core.frame.DataFrame'>

🌒 Print

# zoo.head() # 默认打印前五行
zoo.head(4)
# del zoo['hair'] # 删除 zoo 中 hair 这一列
# zoo.head()

animalClass.head()

🌓 Rows Indexing

# row index location (positional indexing) --> iloc
row = zoo.iloc[0]
# label based indexing --> loc
# row = zoo.loc['bass'] # 相当于用自己命名的 label 替换自带的 index
print(row)

animal_name    aardvark
hair                  1
feathers              0
eggs                  0
milk                  1
airborne              0
aquatic               0
predator              1
toothed               1
backbone              1
breathes              1
venomous              0
fins                  0
legs                  4
tail                  0
domestic              0
catsize               1
class_type            1
Name: 0, dtype: object

🌔 Column Indexing

# according ['column name']
col = zoo['eggs']
# according '.'
col = zoo.eggs
print(col)

0      0
1      0
2      1
3      0
4      0
      ..
96     0
97     1
98     0
99     1
100    1
Name: eggs, Length: 101, dtype: int64

🌕 Describe the Data

# describe each column
zoo.describe()

# describe one column
zoo.eggs.describe()

count    101.000000
mean       0.584158
std        0.495325
min        0.000000
25%        0.000000
50%        1.000000
75%        1.000000
max        1.000000
Name: eggs, dtype: float64

zoo.mean() # .min() .max() .std() .mode() .corr() .mean()
zoo.corr() # 计算相不同列（Series）之间的（DataFrame）
zoo.corrwith(zoo.eggs) # 分别计算 zoo（DataFrame）中所有列与 eggs（Series）的相关系数
zoo.eggs.corr(zoo.milk) # 计算相关系数（one Series with the other Series）

-0.9388478737749797

🌖 Filter the Data

• filter.any() 相当于对 filter 进行或 (or) 运算，寻找 True
• filter.all() 相当于对 filter 进行与 (and) 运算，寻找 False

# choose some rows that satisfy the filter(zoo.eggs==0)
myFilter = zoo.eggs==0 # <1 or >1
print("myFilter:\n",myFilter)
f_all = myFilter.any() # find the True  --> return Ture
f_any = myFilter.all() # find the False --> return False
print("myFilter.any() : %s  \nmyFilter.all() : %s"%(f_all,f_any))

myFilter:
 0       True
1       True
2      False
3       True
4       True
       ...  
96      True
97     False
98      True
99     False
100    False
Name: eggs, Length: 101, dtype: bool
myFilter.any() : True  
myFilter.all() : False

🌗 Clean the Data

zoo.shape # not zoo.shape()

(101, 18)

zoo.head(4)

zoo.isnull().head(4)

zoo.isnull().any().any() # try to find whether zoo has null values（none）

False

zoo.eggs[0] = None # create the null value
zoo.head(4)

zoo = zoo.dropna() # delete the rows containing null values
zoo.head(4)

🌑 Others

append

dfa.append(dfb) 用于将 DataFrame b 连接到 DataFrame a 后面，需要注意的是，这里和 lista.append(listb) 是有去别的，即 pandas 并不在原来的 DataFrame a 上直接操作，而是返回添加好的 DataFrame result 。正确的使用方法应该是 dfa = dfa.append(dfb) 。