File size: 4,818 Bytes
9eaac57
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
from __future__ import annotations

from pathlib import Path
from typing import Tuple
import polars as pl


def build_ticket_views(run_dir: Path) -> Tuple[Path, Path, Path]:
    """Materialize post-run ticket views from events.csv into Parquet files.

    Creates three artifacts in the run directory:
      - ticket_journal.parquet
      - ticket_summary.parquet
      - ticket_state_spans.parquet

    Returns paths to the three files in the order above.
    """
    run_dir = Path(run_dir)
    events_csv = run_dir / "events.csv"
    if not events_csv.exists():
        raise FileNotFoundError(f"events.csv not found in run dir: {events_csv}")

    journal_pq = run_dir / "ticket_journal.parquet"
    summary_pq = run_dir / "ticket_summary.parquet"
    spans_pq = run_dir / "ticket_state_spans.parquet"

    events = pl.scan_csv(str(events_csv))
    # Normalize and order
    journal = (
        events.with_columns(
            [
                pl.col("date").str.to_date().alias("date"),
            ]
        )
        .sort(["case_id", "date"])  # lazy
        .with_columns(
            [
                pl.arange(0, pl.len()).over("case_id").alias("seq_no"),
            ]
        )
        .collect(streaming=True)
    )
    journal.write_parquet(str(journal_pq))

    # Outcomes for counts
    heard = journal.filter(pl.col("type") == "outcome").with_columns(
        (pl.col("detail") == "heard").alias("is_heard")
    )

    base_summary = journal.group_by("case_id").agg(
        [
            pl.first("case_type").alias("case_type"),
            pl.first("date").alias("first_seen_date"),
            pl.last("date").alias("last_seen_date"),
            pl.col("stage").sort_by("seq_no").last().alias("current_stage"),
            (pl.col("type") == "stage_change")
            .cast(pl.Int64)
            .sum()
            .alias("stage_changes"),
            (pl.col("type") == "ripeness_change")
            .cast(pl.Int64)
            .sum()
            .alias("ripeness_transitions"),
        ]
    )
    outcome_summary = (
        heard.group_by("case_id")
        .agg(
            [
                pl.len().alias("total_hearings"),
                pl.col("is_heard").cast(pl.Int64).sum().alias("heard_count"),
            ]
        )
        .with_columns(
            (pl.col("total_hearings") - pl.col("heard_count")).alias("adjourned_count")
        )
    )
    disposed = (
        journal.filter(pl.col("type") == "disposed")
        .group_by("case_id")
        .agg([pl.min("date").alias("disposal_date")])
    )

    summary = (
        base_summary.join(outcome_summary, on="case_id", how="left")
        .with_columns(
            [
                pl.col("total_hearings").fill_null(0),
                pl.col("heard_count").fill_null(0),
                pl.col("adjourned_count").fill_null(0),
            ]
        )
        .join(disposed, on="case_id", how="left")
        .with_columns(
            [
                # Compute age in full days from first to last seen.
                # Use total_days() on duration to be compatible across Polars versions.
                (pl.col("last_seen_date") - pl.col("first_seen_date"))
                .dt.total_days()
                .alias("age_days_end"),
                pl.when(pl.col("disposal_date").is_not_null())
                .then(pl.lit("DISPOSED"))
                .otherwise(pl.lit("ACTIVE"))
                .alias("final_status"),
            ]
        )
    )
    summary.write_parquet(str(summary_pq))

    # Spans from stage changes
    sc = (
        journal.filter(pl.col("type") == "stage_change")
        .select(["case_id", "date", "stage"])
        .rename({"date": "start_date"})
    )
    spans = sc.with_columns(
        [
            pl.col("start_date").shift(-1).over("case_id").alias("end_date"),
        ]
    ).with_columns(
        [
            pl.when(pl.col("end_date").is_null())
            .then(pl.col("start_date"))
            .otherwise(pl.col("end_date"))
            .alias("end_date")
        ]
    )
    spans.write_parquet(str(spans_pq))
    return journal_pq, summary_pq, spans_pq


def load_ticket_views(run_dir: Path):
    """Load ticket views; build them if missing. Returns (journal, summary, spans).

    Uses Polars DataFrames if Polars is available; otherwise returns pandas DataFrames.
    """
    run_dir = Path(run_dir)
    journal_pq = run_dir / "ticket_journal.parquet"
    summary_pq = run_dir / "ticket_summary.parquet"
    spans_pq = run_dir / "ticket_state_spans.parquet"

    if not (journal_pq.exists() and summary_pq.exists() and spans_pq.exists()):
        build_ticket_views(run_dir)

    journal = pl.read_parquet(str(journal_pq))
    summary = pl.read_parquet(str(summary_pq))
    spans = pl.read_parquet(str(spans_pq))
    return journal, summary, spans