擅长:python、mysql、java
<h3><a href="https://pandas.pydata.org/pandas-docs/stable/generated/pandas.DataFrame.groupby.html" rel="nofollow noreferrer">^{<cd1>}</a>+<a href="https://pandas.pydata.org/pandas-docs/stable/generated/pandas.Series.cumsum.html" rel="nofollow noreferrer">^{<cd2>}</a>+<a href="https://pandas.pydata.org/pandas-docs/stable/generated/pandas.Series.value_counts.html" rel="nofollow noreferrer">^{<cd3>}</a></h3>
<p>可以将<code>groupby</code>与自定义函数一起使用:</p>
<pre><code>df = pd.DataFrame({'ID':[400, 400, 400, 400, 400, 400, 500, 500, 500, 500],
'Number':[1, 2, 3, 4, 8, 9, 22, 23, 26, 27]})
def mean_count(x):
return (x - x.shift()).ne(1).cumsum().value_counts().mean()
res = df.groupby('ID')['Number'].apply(mean_count).reset_index()
print(res)
ID Number
0 400 3.0
1 500 2.0
</code></pre>