组中列的具体值的计数

Question

我正在努力实现以下目标：

aggregate_commitment_df = commitment_df.group_by("sprint_start_date", "squad").agg(
    stories_committed=pl.count(pl.col("issue_type") == "Story"),
    spikes_committed=pl.count(pl.col("issue_type") == "Spike"),
    bugs_committed=pl.count(pl.col("issue_type") == "Bug"),
    story_points_committed=pl.sum("story_points"),
)

，例如，

stories_committed

将是每个

issue_type == "Story"

和

sprint_start_date

具有

squad

的所有行的计数。

Answer 1

在使用 count 之前，您需要在聚合上下文中使用

filter

。

import polars as pl
from datetime import date

commitment_df = pl.DataFrame({
    "sprint_start_date": [date(2023,8,1),date(2023,8,2),date(2023,8,3),] * 8,
    "squad": ["team1", "team2"] * 12,
    "issue_type": ["Story"] * 10 + ["Spike"] * 8 + ["Bug"] * 6, 
    "story_points": [*range(4)] * 6,
})

aggregate_commitment_df = commitment_df.group_by("sprint_start_date", "squad").agg(
    stories_committed=pl.col("issue_type").filter(pl.col("issue_type") == "Story").count(),
    spikes_committed=pl.col("issue_type").filter(pl.col("issue_type") == "Spike").count(),
    bugs_committed=pl.col("issue_type").filter(pl.col("issue_type") == "Bug").count(),
    story_points_committed=pl.sum("story_points"),
)

print(aggregate_commitment_df)

结果：

shape: (6, 6)
┌───────────────────┬───────┬───────────────────┬──────────────────┬────────────────┬────────────────────────┐
│ sprint_start_date ┆ squad ┆ stories_committed ┆ spikes_committed ┆ bugs_committed ┆ story_points_committed │
│ ---               ┆ ---   ┆ ---               ┆ ---              ┆ ---            ┆ ---                    │
│ date              ┆ str   ┆ u32               ┆ u32              ┆ u32            ┆ i64                    │
╞═══════════════════╪═══════╪═══════════════════╪══════════════════╪════════════════╪════════════════════════╡
│ 2023-08-03        ┆ team2 ┆ 1                 ┆ 2                ┆ 1              ┆ 8                      │
│ 2023-08-02        ┆ team1 ┆ 1                 ┆ 2                ┆ 1              ┆ 4                      │
│ 2023-08-01        ┆ team1 ┆ 2                 ┆ 1                ┆ 1              ┆ 4                      │
│ 2023-08-03        ┆ team1 ┆ 2                 ┆ 1                ┆ 1              ┆ 4                      │
│ 2023-08-02        ┆ team2 ┆ 2                 ┆ 1                ┆ 1              ┆ 8                      │
│ 2023-08-01        ┆ team2 ┆ 2                 ┆ 1                ┆ 1              ┆ 8                      │
└───────────────────┴───────┴───────────────────┴──────────────────┴────────────────┴────────────────────────┘