Series type data of Pandas series

Posted by vexx on Thu, 30 Dec 2021 00:40:20 +0100

Series type data of Pandas series

This article begins to write a series of articles on Pandas, starting with: how to create data in Pandas. There are two types of data created in Pandas:

  • Series type
  • DataFrame type

<!--MORE-->

Content map

Series type

Series is a one-dimensional array structure, which is only composed of index and value.

Series indexes are unique. Indexes can be either numbers or characters. The system will automatically convert them into an object type (character type in pandas).

DataFrame type

DataFrame is a two-dimensional data structure that combines several Series by column. Each column is taken out separately as a Series; Besides index and value, there is also column. In the following figure:

  • Index: 0, 1, 2, 3
  • Field attribute: fruit, number
  • value: apple, grape, etc; 200, 300, etc

Import library

Import two libraries first:

import pandas as pd
import numpy as np

Series type creation and operation

  • Generated by iteratable type list and tuple
  • Generated by python dictionary
  • Generated by numpy array

List generation

Generate Series data by list

s1 = pd.Series([7,8,9,10])
s1

# result
0     7
1     8
2     9
3    10
dtype: int64
s2 = pd.Series(list(range(1,8)))
s2

# result
0    1
1    2
2    3
3    4
4    5
5    6
6    7
dtype: int64

Tuple generation

The following method is to generate Series data through tuples

s3 = pd.Series((7,8,9,10,11))
s3

# result
0     7
1     8
2     9
3    10
4    11
dtype: int64

s4 = pd.Series(tuple(range(1,8)))  #  From 1 to 8, excluding 8
s4

# result
0    1
1    2
2    3
3    4
4    5
5    6
6    7
dtype: int64

Create using dictionary

The key of the dictionary is the index, and the value is the value corresponding to the Series structure

dic_data = {"0":"Apple", "1":"Banana", "2":"Hami melon","3":"orange"}

s5 = pd.Series(dic_data)
s5

# result
0     Apple
1     Banana
2    Hami melon
3     orange
dtype: object

Using numpy arrays

s6 = pd.Series(np.arange(3,9))
s6

# result
0    3
1    4
2    5
3    6
4    7
5    8
dtype: int64

Specify index (list)

The default indexes are numeric values starting from 0. You can specify each index when creating

# default

s1 = pd.Series([7,8,9,10])
s1

# result
0     7
1     8
2     9
3    10
dtype: int64
s7 = pd.Series([7,8,9,10], index=["A","B","C","D"])  # Specify index value 
s7

# result
A     7
B     8
C     9
D    10
dtype: int64

Specify index (dictionary form)

Dictionary key as index value

dic_data = {"Fruit 1":"Apple", 
            "Fruit 2":"Banana", 
            "Fruit 3":"Hami melon",
            "Fruit 4":"orange"
           }

s8 = pd.Series(dic_data)
s8

# result
 Fruit 1     Apple
 Fruit 2     Banana
 Fruit 3    Hami melon
 Fruit 4     orange
dtype: object

View index values

s8

# result
 Fruit 1     Apple
 Fruit 2     Banana
 Fruit 3    Hami melon
 Fruit 4     orange
dtype: object
s8.index   # View index values

# result
Index(['Fruit 1', 'Fruit 2', 'Fruit 3', 'Fruit 4'], dtype='object')

View values

s8

# result
 Fruit 1     Apple
 Fruit 2     Banana
 Fruit 3    Hami melon
 Fruit 4     orange
dtype: object
s8.values

# result
array(['Apple', 'Banana', 'Hami melon', 'orange'], dtype=object)

Change index

# 1. New index
index_new = ['one', 'two', 'three', 'four'] 

# 2. Assignment
s8.index = index_new

s8
# result
one       Apple
two       Banana
three    Hami melon
four      orange
dtype: object

Check for null values

s7

# result
A     7
B     8
C     9
D    10
dtype: int64
s7.isnull()  # No null value

# result
A    False
B    False
C    False
D    False
dtype: bool
s7.notnull()

# result
A    True
B    True
C    True
D    True
dtype: bool

View the value of an index

s7

A     7
B     8
C     9
D    10
dtype: int64

There are two ways to view:

  • View by custom index
  • View through the corresponding numerical index
s7["A"]  #  Custom index value

7
s7[0]   # Default numeric index

7
s7["D"]

10
s7[3]

10

Convert Series to dictionary

s_dic = s7.to_dict()  # Convert to dictionary form
s_dic

# result
{'A': 7, 'B': 8, 'C': 9, 'D': 10}
type(s_dic)   # The result is displayed as a dictionary type

# result
dict

Name the Series index

s8

# result
one       Apple
two       Banana
three    Hami melon
four      orange
dtype: object
s8.index  # Original index

Index(['one', 'two', 'three', 'four'], dtype='object')
s8.index.name = "Fruits"  # Index naming
s8

The results are displayed as:

Fruits
one       Apple
two       Banana
three    Hami melon
four      orange
dtype: object


s8.index   # Index after change
Index(['one', 'two', 'three', 'four'], dtype='object', name='Fruits')

Modify Series values

s8

# The result is
 Fruits
one       Apple
two       Banana
three    Hami melon
four      orange
dtype: object
s8["three"] = "watermelon"  # Equivalent to s8[2] = "watermelon"

s8

The changed value is:

Fruits
one      Apple
two      Banana
three    watermelon
four     orange
dtype: object

Convert Series structure to DataFrame structure

s8

Fruits
one      Apple
two      Banana
three    watermelon
four     orange
dtype: object

Three functions are involved in the process of converting s8 to DataFrame:

  • to_frame: convert to DataFrame
  • reset_index: index reset of DataFrame type
  • rename: reset the field properties of the DataFrame

Please look forward to the detailed explanation of DataFrame in the next section!

Extended reading

Many knowledge points of pandas are used in the previously written travel strategy articles for learning:

Topics: Data Analysis

Hot Topics

©2024 Programmer Think