Python 3 Data Science Summary: https://blog.csdn.net/weixin_41793113/article/details/99707225 

 

import numpy as np
import pandas as pd
from pandas import Series, DataFrame

data = {
    'Country': ['China', 'India', 'Brazil'],
    'Capital': ['Beijing', 'New Delhi', 'Brasilia'],
    'Population': ['1432732201', '1303171635', '207847528']
} #python dictionary

s1 = Series(data['Country'])

s1

s2 = Series(data['Capital'])
s3 = Series(data['Population'])

df = DataFrame(data) #DataFrame can be understood as a multi-column Series

df

for row in df.iterrows():
    print(row) #Iterator traverses each line

for row in df.iterrows():
    print(row[0], '+', row[1]) #[0] is the index [1] is the value
    break

# Creating DataFrame through Series

df_new = DataFrame([s1,s2,s3])

df_new

df #The difference lies in listing.

df_new = DataFrame([s1,s2,s3], index=['Country','Capital', 'Population'])
#Solve the above column name problem and pass in the index

df_new

# It can be found that it is superimposed by line and needs to be transposed

df_new.T

 

# Detailed description of the basic attributes of Pandas DataFrame

List of Basic Functions

import pandas as pd import library

df = pd.DataFrame(data=None, index=None, columns=None, dtype=None, copy=False)
Create a DataFrame

	Code	function
1	DataFrame()	Create a DataFrame object
2	df.values	Returns an object of type ndarray
3	df.index	Get row index
4	df.columns	Get column index
5	df.axes	Get row and column indexes
6	df.T	Row-column alignment
7	df. info()	Print the information of the DataFrame object
8	df.head(i)	Display the first i-line data
9	df.tail(i)	Display post-i row data
10	df.describe()	View column-by-column statistics of data

 

1. Create a DataFrame

The value of the parameter index of the DataFrame() function corresponds to the value of the row index. If it is not assigned manually, it will be allocated from 0 by default. columns are equivalent to column indexes, and if you do not assign them manually, they will be assigned from 0 by default.

data = {
        'Gender':['male','female','female','male','male'],
        'Full name':['Xiao Ming','Xiaohong','Xiaofang','Big black','Zhang San'],
        'Age':[20,21,25,24,29]}
df = pd.DataFrame(data,index=['one','two','three','four','five'],
               columns=['Full name','Gender','Age','Occupation'])
df

 

Operation results:

2. df.values returns objects of type ndarray

The ndarray type, numpy's N-dimensional array object, usually converts data of DataFrame type to ndarray type, which is easy to operate. For example, slicing a DataFrame type requires the form of df. iloc [:, 1:3], and directly X [:, 1:3] for the array type.

X = df.values
print(type(X)) #Display data type
X

 

3. df.index to get row index

df.index

 

Operation results:

Index(['one', 'two', 'three', 'four', 'five'], dtype='object')

 

4. df.columns retrieves column indexes

df.columns

 

Operation results:

Index(['Full name', 'Gender', 'Age', 'Occupation'], dtype='object')

 

5. df.axes retrieves row and column indexes

df.axes

 

Operation results:

[Index(['one', 'two', 'three', 'four', 'five'], dtype='object'),
 Index(['Full name', 'Gender', 'Age', 'Occupation'], dtype='object')]

• 1
• 2

6. df.T index and columns alignment

df.T

 

Operation results:

7. df.info () Print the information of the DataFrame object

df.info()

 

Operation results:

<class 'pandas.core.frame.DataFrame'>
Index: 5 entries, one to five
Data columns (total 4 columns):
//Name 5 non-null object
//Gender 5 non-null object
//Age 5 non-null int64
//Occupation 0 non-null object
dtypes: int64(1), object(3)
memory usage: 200.0+ bytes

 

8.df.head(i) displays the first i-line data

df.head(2)

 

Operation results:

If you want to display the first few columns of data, you can use df.T.head(i)

9. df.tail(i) post-display i-line data

df.tail(2)

Operation results:

10. df.describe() view statistics by column

Information such as number of data, missing value, minimum and maximum number, average value, quantile can be displayed.