繁体   English   中英

根据 pandas 中另一列的值计算列中 integer 值的数量

[英]Counting the number of integer values in a column depending on the value of another column in pandas

Hi i have a column which has values in both integer and text, I am trying to write a function in python where i can get the count of only integer values corresponding to another value in another column.

听到是数据的样本,

constructorId   positionText
 1                 3
 1                 4
 1                 R
 4                 6
 4                 5
 4                 N
 4                 9

我希望结果是这样的

 constructorID    positionText_count
    1                   2
    4                   3

这是代码

  def not_finished(c):
   r = 0
   for c in hybrid_era_results['constructorId']:
       y = hybrid_era_results['positionText']
       if isinstance(y, int):
           r = r+1
   return r 

此代码不会引发错误,但是当我调用 function 时,它总是返回值 0。我做错了什么?

IIUC,您可以使用to_numeric过滤数值,然后groupby.sum

(pd.to_numeric(df['positionText'], errors='coerce')
   .notna()
   .groupby(df['constructorId'])
   .sum()
   .reset_index()
)

output:

   constructorId  positionText
0              1             2
1              4             3

要设置自定义列名,请使用.agg(positionText_count='sum')代替.sum() (如果需要特殊字符的空格,请使用.agg(**{'positionText_count': 'sum'}) ) :

   constructorId  positionText_count
0              1                   2
1              4                   3

由于没有关于数据的明确细节,这里有一个简单的例子来检查 col2 中 col1 的整数值是否存在,其中两列都是 str 类型。

import pandas as pd

# dimple example df
df = pd.DataFrame({"col1": ["value1", "4",  "5", "3"], "col2": ["3", "6", "value2", "4"]})

# get the index where we have integer in both columns
idx_col1 = df.col1.str.isdigit()
idx_col2 = df.col2.str.isdigit()

# retrieve integer values
df_col1_int = df.loc[idx_col1, "col1"]
df_col2_int = df.loc[idx_col2, "col2"]

# get only values from col1 tht exists in col2 (the same index is not required)    
idx_exists = df_col1_int.isin(df_col2_int)
df_exists = df_col1_int[idx_exists]

# get the number of integers in common
len(df_exists)

暂无
暂无

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM