如何用升序数字前缀替换极性中的空值?

df = df.with_columns(
    pl.col("asset_number").fill_null(
        "PREFIX - " + pl.int_range(pl.len()).cast(pl.String)
    )
)

这里有一种使用 ordinal 排名超过 asset_number 。要添加前导0,我们使用 .str.pad_start .

df.with_columns(
    pl.coalesce(
        "asset_number",
        "PREFIX - "
        + pl.all()
        .rank("ordinal")
        .over("asset_number")
        .cast(pl.String)
        .str.pad_start(3, "0"),
    )
)

输出:

shape: (8, 2)
┌──────────────┬──────────────┐
│ name         ┆ asset_number │
│ ---          ┆ ---          │
│ str          ┆ str          │
╞══════════════╪══════════════╡
│ Office Chair ┆ PREFIX - 001 │
│ Office Chair ┆ PREFIX - 002 │
│ Office Chair ┆ PREFIX - 003 │
│ Office Chair ┆ CMP - 001    │
│ Office Chair ┆ CMP - 005    │
│ Office Chair ┆ PREFIX - 004 │
│ Table        ┆ PREFIX - 005 │
│ Table        ┆ CMP - 007    │
└──────────────┴──────────────┘

完整代码:

import polars as pl

df = pl.DataFrame(
    {
        "name": [
            "Office Chair",
            "Office Chair",
            "Office Chair",
            "Office Chair",
            "Office Chair",
            "Office Chair",
            "Table",
            "Table",
        ],
        "asset_number": [
            None,
            None,
            None,
            "CMP - 001",
            "CMP - 005",
            None,
            None,
            "CMP - 007",
        ],
    }
)

df = df.with_columns(
    pl.coalesce(
        "asset_number",
        "PREFIX - "
        + pl.all()
        .rank("ordinal")
        .over("asset_number")
        .cast(pl.String)
        .str.pad_start(3, "0"),
    )
)

print(df)