[英]Pandas: Incrementally count occurrences in a column
I have a DataFrame (df) which contains a 'Name' column. 我有一个DataFrame(df),其中包含一个'Name'列。 In a column labeled 'Occ_Number' I would like to keep a running tally on the number of appearances of each value in 'Name'. 在标有“Occ_Number”的列中,我想保持“名称”中每个值的出现次数的运行记录。
For example: 例如:
Name Occ_Number
abc 1
def 1
ghi 1
abc 2
abc 3
def 2
jkl 1
jkl 2
I've been trying to come up with a method using 我一直试图想出一个使用的方法
>df['Name'].value_counts()
but can't quite figure out how to tie it all together. 但无法弄清楚如何将它们联系在一起。 I can only get the grand total from value_counts(). 我只能从value_counts()获得总计。 My process thus far involves creating a list of the 'Name' column string values which contain counts greater than 1 with the following code: 到目前为止,我的过程涉及使用以下代码创建包含大于1的计数的“名称”列字符串值的列表:
>things = df['Name'].value_counts()
>things = things[things > 1]
>queries = things.index.values
I was hoping to then somehow cycle through 'Name' and conditionally add to Occ_Number by checking against queries, but this is where I'm getting stuck. 我希望以某种方式循环“名称”并通过检查查询有条件地添加到Occ_Number,但这是我被卡住的地方。 Does anybody know of a way to do this? 有人知道这样做的方法吗? I would appreciate any help. 我将不胜感激任何帮助。 Thank you! 谢谢!
You can add a helper column and then use cumsum
: 您可以添加辅助列,然后使用cumsum
:
df =pd.DataFrame({'Name':['abc', 'def', 'ghi', 'abc', 'abc', 'def', 'jkl', 'jkl']})
add count: 添加计数:
df['counts'] =1
group by name: 按名称分组:
cs =df.groupby('Name')['counts'].cumsum()
# set series name
cs.name = 'Occ_number'
join series back to dataframe: 将系列连接回数据帧:
# remove helper column
del df['counts']
df.join(cs)
returns: 收益:
Name Occ_number
0 abc 1
1 def 1
2 ghi 1
3 abc 2
4 abc 3
5 def 2
6 jkl 1
7 jkl 2
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.