如何在Polars中筛选任意列包含特定值的所有行（附Pandas实现对比）

Hi there! Since you already know how to do this in Pandas, translating that logic to Polars is straightforward—here's how you can achieve the same result:

Step-by-Step Implementation #

First, let's start with your existing Polars DataFrame, then apply the filter:

import polars as pl
df = pl.from_repr("""
┌─────┬─────┬─────┐
│ a   ┆ b   ┆ c   │
│ --- ┆ --- ┆ --- │
│ str ┆ str ┆ str │
╞═════╪═════╪═════╡
│ Yes ┆ No  ┆ No  │
│ No  ┆ No  ┆ No  │
│ Yes ┆ No  ┆ Yes │
└─────┴─────┴─────┘
""")

# Filter rows where at least one column equals "Yes"
filtered_df = df.filter(pl.any_horizontal(pl.all().eq("Yes")))
print(filtered_df)

What This Code Does #

Let's break it down to match your Pandas logic:

• pl.all().eq("Yes"): Checks every column to see if each element is equal to "Yes", returning a boolean DataFrame (just like df_pd.eq('Yes') in Pandas).
• pl.any_horizontal(): Runs a row-wise OR operation on those boolean values—this is the equivalent of any(axis=1) in Pandas, returning True for any row that has at least one "Yes".
• .filter(): Uses those boolean results to keep only the rows that match the condition, just like df_pd.loc[...].

Output #

Running the code will give you exactly the result you want:

┌─────┬─────┬─────┐
│ a   ┆ b   ┆ c   │
│ --- ┆ --- ┆ --- │
│ str ┆ str ┆ str │
╞═════╪═════╪═════╡
│ Yes ┆ No  ┆ No  │
│ Yes ┆ No  ┆ Yes │
└─────┴─────┴─────┘

Optional: Explicit Column Selection #

If you prefer to specify columns explicitly (instead of using pl.all()), you can write it like this too—great for clarity when working with larger datasets:

filtered_df = df.filter(pl.any_horizontal(pl.col("a", "b", "c").eq("Yes")))

内容的提问来源于stack exchange，提问作者nickanor