Pandas GroupBy – 计算每个组合的出现次数

在本文中，我们将 GroupBy 两列并计算 Pandas 中每个组合的出现次数。

DataFrame.groupby() 方法用于将 DataFrame 分成组。它将生成数据框特定列中存在的类似数据计数的数量。

Syntax: DataFrame.groupby(by=None, axis=0, level=None )

Parameters:

• by: mapping, function, string, label, or iterable to group elements.
• axis : group by along with the row (axis=0) or column (axis=1).
• level: Integer. value to the group by a particular level or levels.

为了理解这个概念，我们将使用下面给出的简单数据集：

Python3

# Import library
import pandas as pd
import numpy as np
 
# initialise data of lists.
Data = {'Products':['Box','Color','Pencil','Eraser','Color',
                    'Pencil','Eraser','Color','Color','Eraser','Eraser','Pencil'],
 
       'States':['Jammu','Kolkata','Bihar','Gujrat','Kolkata',
                 'Bihar','Jammu','Bihar','Gujrat','Jammu','Kolkata','Bihar'],
 
       'Sale':[14,24,31,12,13,7,9,31,18,16,18,14]}
 
# Create DataFrame
df = pd.DataFrame(Data, columns=['Products','States','Sale'])
 
# Display the Output
display(df)

输出：

方法一：使用 Pandas dataframe.size()

它返回元素的总数，通过将 shape 方法返回的行和列相乘进行比较。

Syntax: dataframe.size

蟒蛇3

new = df.groupby(['States','Products']).size()
display(new)

输出：

方法 2：使用 Pandas dataframe.count()

它用于计算编号。给定轴上的非 NA/null 观测值。它也适用于非浮动类型的数据。

Syntax: DataFrame.count(axis=0, level=None, numeric_only=False)

Parameters:

• axis : 0 or ‘index’ for row-wise, 1 or ‘columns’ for column-wise
• level : If the axis is a MultiIndex (hierarchical), count along a particular level, collapsing into a DataFrame
• numeric_only : Include only float, int, boolean data

Returns: count : Series (or DataFrame if level specified)

蟒蛇3

new = df.groupby(['States','Products'])['Sale'].count()
display(new)

输出：

方法 3：使用 Pandas reset_index()

它是一种重置数据帧索引的方法。reset_index() 方法将范围从 0 到数据长度的整数列表设置为索引。

Syntax: DataFrame.reset_index(level=None, drop=False, inplace=False, col_level=0, col_fill=”)

Parameters:

• level: int, string or a list to select and remove passed column from index.
• drop: Boolean value, Adds the replaced index column to the data if False.
• inplace: Boolean value, make changes in the original data frame itself if True.
• col_level: Select in which column level to insert the labels.
• col_fill: Object, to determine how the other levels are named.

Return type: DataFrame

蟒蛇3

new = df.groupby(['States','Products'])['Sale'].agg('count').reset_index()
display(new)

输出：

方法 4：使用 pandas.pivot()函数

它基于 DataFrame 的 3 列生成数据透视表。使用索引/列中的唯一值并填充值。

Syntax: pandas.pivot(index, columns, values)

Parameters:

• index[ndarray] : Labels to use to make new frame’s index
• columns[ndarray] : Labels to use to make new frame’s columns
• values[ndarray] : Values to use for populating new frame’s values

Returns: Reshaped DataFrame
Exception: ValueError raised if there are any duplicates.

蟒蛇3

new = df.groupby(['States','Products'],as_index = False
                ).count().pivot('States','Products').fillna(0)
display(new)

输出：