擅长:python、mysql、java
<p>使用<a href="https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.filter.html" rel="nofollow noreferrer">^{<cd1>}</a>过滤包含<code>rater</code>等列的数据框,然后使用<a href="https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.ne.html" rel="nofollow noreferrer">^{<cd3>}</a>沿<code>axis=0</code>比较包含<code>rater</code>的列与<code>right_answer</code>列,然后使用<a href="https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.sum.html" rel="nofollow noreferrer">^{<cd7>}</a>沿<code>axis=1</code>获取给出错误答案的<code>raters</code>数,然后使用<a href="https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.Series.ge.html" rel="nofollow noreferrer">^{<cd10>}</a>创建布尔掩码,最后使用此筛选数据框行<code>mask</code>:</p>
<pre><code>mask = (
df.filter(like='rater')
.ne(df['right_answer'], axis=0).sum(axis=1).ge(2)
)
df = df[mask]
</code></pre>
<p>结果:</p>
<pre><code># print(df)
right_answer rater1 rater2 rater3 item
1 1 1 2 2 S02
2 2 1 2 1 S03
</code></pre>