如何按计数值分组（python，pandas）

huangapple
go评论93阅读模式
英文:

How to group by counted value (python, pandas)

问题

你好，你可以尝试使用以下代码来实现你的需求：

  1. import pandas as pd
  3. # 创建DataFrame
  4. data = {'SerialNr': [1, 2, 3, 4],
  5.         'Amount': [2, 4, 3, 2]}
  6. df = pd.DataFrame(data)
  8. # 使用groupby和count函数来计算数量
  9. result = df.groupby('Amount')['SerialNr'].count().reset_index()
  10. result.columns = ['Amount', 'Quantity of SerialNr']
  12. # 打印结果
  13. print(result)

这段代码会根据 "Amount" 列进行分组，然后计算每个分组中的 "SerialNr" 数量，最后得到你所需的结果表格。

英文:

Hello I have this Dataframe:

SerialNr Amount
1 2
2 4
3 3
4 2

And I want a table that shows me how many serial numbers in the right column have the same numbers.
Here for example it would look like this

Amount Quantity of SerialNr
1 0
2 2
3 1
4 1

It should sum up the Serialnummer with the same amount

Tried to do it with groupby but did not work

答案1

得分: 1

根据您的期望输出需要Amount的数量，因此可以使用Series.value_counts结合Series.reindex，以1为起始点进行如下操作：

  1. s = df['Amount'].value_counts()
  3. df = (s.reindex(range(1, s.index.max()+1), fill_value=0)
  4.        .rename_axis('Amount')
  5.        .reset_index(name='Amount的数量'))
  6. print (df)
  7.    Amount  Amount的数量
  8. 0       1           0
  9. 1       2           2
  10. 2       3           1
  11. 3       4           1

或者使用Categorical：

  1. cats = range(1, df.Amount.max()+1)
  2. df = (pd.Categorical(df['Amount'], categories=cats)
  3.         .value_counts()
  4.         .rename_axis('Amount')
  5.         .reset_index(name='Amount的数量'))
  7. print (df)
  8.   Amount  Amount的数量
  9. 0      1           0
  10. 1      2           2
  11. 2      3           1
  12. 3      4           1
英文:

From your expected ouput need Quantity of Amount, so use Series.value_counts with Series.reindex by range starting by 1:

  1. s = df['Amount'].value_counts()
  3. df = (s.reindex(range(1, s.index.max()+1), fill_value=0)
  4.        .rename_axis('Amount')
  5.        .reset_index(name='Quantity of Amount'))
  6. print (df)
  7.    Amount  Quantity of SerialNr
  8. 0       1                     0
  9. 1       2                     2
  10. 2       3                     1
  11. 3       4                     1

Or use Categorical:

  1. cats = range(1, df.Amount.max()+1)
  2. df = (pd.Categorical(df['Amount'], categories=cats)
  3.         .value_counts()
  4.         .rename_axis('Amount')
  5.         .reset_index(name='Quantity of Amount'))
  7. print (df)
  8.   Amount  Quantity of Amount
  9. 0      1                   0
  10. 1      2                   2
  11. 2      3                   1
  12. 3      4                   1

答案2

得分: 1

另一个可能的解决方案：

  1. (pd.concat([df.set_index('SerialNr'), 
  2.             df['Amount'].value_counts().rename('Amount的数量')], axis=1)
  3.  .fillna(0, downcast='infer').reset_index()
  4.  .drop('Amount', axis=1).rename({'index': 'Amount'}, axis=1))

输出：

  1.    Amount  count
  2. 0       1      0
  3. 1       2      2
  4. 2       3      1
  5. 3       4      1
英文:

Another possible solution:

  1. (pd.concat([df.set_index('SerialNr'), 
  2.             df['Amount'].value_counts().rename('Quantity of Amount')], axis=1)
  3.  .fillna(0, downcast='infer').reset_index()
  4.  .drop('Amount', axis=1).rename({'index': 'Amount'}, axis=1))

Output:

  1.    Amount  count
  2. 0       1      0
  3. 1       2      2
  4. 2       3      1
  5. 3       4      1

huangapple
  • 本文由 发表于 2023年6月26日 17:42:59
  • 转载请务必保留本文链接：https://go.coder-hub.com/76555465.html
  • pandas
  • python
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定