Pandas Library common functions

• One 、 Select rows and columns
• • 1.1 Create sample data
  • 1.1 Select the row that meets the condition according to the condition
  • 1.2 Use loc Make a selection
  • • 1.2.1 Select row
    • 1.2.2 Select column
  • 1.3 Use iloc Make a selection
  • • 1.3.1 Select row
    • 1.3.2 Select column
  • 1.4 Use it directly column Select column
  • 1.5 Select a piece of data
• Two 、 duplicate removal

One 、 Select rows and columns

1.1 Create sample data

import pandas as pd
import numpy as np
value = np.array([[1, 2, 3, 4], [2, 3, 4, 5], [3, 4, 5, 6]])
index = np.array([10, 20, 30])
column = np.array(['u', 'i', 'r', 'time'])
df = pd.DataFrame(data=value, index=index, columns=column)
''' u i r time 10 1 2 3 4 20 2 3 4 5 30 3 4 5 6 '''

Mainly involves loc and iloc Usage of , The old version of python There's also ix, however ix Has been deprecated by the new edition

.ix is deprecated. Please use
.loc for label based indexing or
.iloc for positional indexing

It's easy to see ,loc Index by attribute name , and iloc Is indexed by column number , Such as df.loc[:, ‘u’] and df.iloc[:, 0] It implements the same function ~

1.1 Select the row that meets the condition according to the condition

# Filter through a Boolean decision , select time Column greater than 4 The line of 
''' u i r time 20 2 3 4 5 30 3 4 5 6 '''
data = df[df.time > 4]
# You can also use to set multiple conditions , If the requirements are added, the score will be 5 The conditions of division 
''' u i r time 30 3 4 5 6 '''
data = df[(df.time > 4) & (df.r == 5)]

1.2 Use loc Make a selection

1.2.1 Select row

# Choose the first line （ First line index yes 10）
''' u 1 i 2 r 3 time 4 Name: 10, dtype: int32 '''
data = df.loc[10]
# Select the second and subsequent rows （ Second elements index yes 20）
''' u i r time 20 2 3 4 5 30 3 4 5 6 '''
data = df.loc[20:30]
# perhaps data = df.loc[20:]

1.2.2 Select column

# Choose the first column （ In the first column column yes 'u'）
''' u 1 i 2 r 3 time 4 Name: 10, dtype: int32 '''
data = df.loc[:, 'u']
# Select the second to fourth columns （ The second column column yes 'i', In the fourth column column yes ‘time’）
''' i r time 10 2 3 4 20 3 4 5 30 4 5 6 '''
data = df.loc[:, 'i':'time']

1.3 Use iloc Make a selection

1.3.1 Select row

# Choose the first line 
''' u 1 i 2 r 3 time 4 Name: 10, dtype: int32 '''
data = df.iloc[0]
# Select the second and subsequent rows 
''' u i r time 20 2 3 4 5 30 3 4 5 6 '''
data = df.iloc[1:3]
# perhaps data = df.iloc[1:]

1.3.2 Select column

# Choose the first column 
''' u 1 i 2 r 3 time 4 Name: 10, dtype: int32 '''
data = df.iloc[:, 0]
# Select the second to fourth columns 
''' i r time 10 2 3 4 20 3 4 5 30 4 5 6 '''
data = df.iloc[:, 1:4]

1.4 Use it directly column Select column

# Choose the first column （ In the first column column yes 'u'）
''' u 1 i 2 r 3 time 4 Name: 10, dtype: int32 '''
data = df['u']

1.5 Select a piece of data

# Use loc
''' u i r 10 1 2 3 20 2 3 4 '''
data = df.loc[10:20, 'u':'r']
# Use iloc
''' u i r 10 1 2 3 20 2 3 4 '''
data = df.iloc[0:2, 0:3]

Two 、 duplicate removal

import pandas as pd
# There are four columns of data , The names are u,i,r and time
#u i r time
#1 3 4 1
#2 1 5 2
#3 1 5 3
#1 3 4 2
#1 3 4 1
df = pd.read_csv('rating.txt', names=['u', 'i', 'r', 'time'])

Obviously , The first data is the same as the fifth data , Data sets need to be de duplicated , The main use is drop_duplicates()

# u i r time
#0 1 3 4 1
#1 2 1 5 2
#2 3 1 5 3
#3 1 3 4 2
data = df.drop_duplicates()

But if you don't think about time , We will find the first 、 Four 、 Five pieces of data are repeated , You can combine the above method of selecting columns to remove duplicates

# u i r
#0 1 3 4
#1 2 1 5
#2 3 1 5
data = df.loc[:, ['u', 'i', 'r']].drop_duplicates()