'''
PM4Py – A Process Mining Library for Python
Copyright (C) 2024 Process Intelligence Solutions UG (haftungsbeschränkt)
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU Affero General Public License as
published by the Free Software Foundation, either version 3 of the
License, or any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU Affero General Public License for more details.
You should have received a copy of the GNU Affero General Public License
along with this program. If not, see this software project's root or
visit <https://www.gnu.org/licenses/>.
Website: https://processintelligence.solutions
Contact: info@processintelligence.solutions
'''
"""Utilities for enriching event data stored in Polars LazyFrames."""
from enum import Enum
from typing import Optional, Dict, Any, Iterable, List, Set
import polars as pl
from pm4py.util import constants, xes_constants, exec_utils
[docs]
class Parameters(Enum):
CASE_ID_KEY = constants.PARAMETER_CONSTANT_CASEID_KEY
START_TIMESTAMP_KEY = constants.PARAMETER_CONSTANT_START_TIMESTAMP_KEY
TIMESTAMP_KEY = constants.PARAMETER_CONSTANT_TIMESTAMP_KEY
COMPUTE_EXTRA_TEMPORAL_FEATURES = "compute_extra_temporal_features"
def _drop_if_present(lf: pl.LazyFrame, cols: Iterable[str]) -> pl.LazyFrame:
"""Drop columns from a LazyFrame if they exist (schema-aware)."""
existing: Set[str] = set(lf.collect_schema().names())
to_drop = [c for c in cols if c in existing]
return lf.drop(to_drop) if to_drop else lf
def _prepare_case_features(
df: pl.LazyFrame,
case_id_key: str,
start_timestamp_key: str,
timestamp_key: str,
compute_extra_temporal_features: bool,
) -> pl.LazyFrame:
"""Compute per-case aggregates needed for feature enrichment."""
# Use more specific internal names to minimize collisions with user columns.
case_start_col = "__pm4py_case_start"
case_end_col = "__pm4py_case_end"
case_summary = df.group_by(case_id_key).agg(
[
pl.col(start_timestamp_key).first().alias(case_start_col),
pl.col(timestamp_key).last().alias(case_end_col),
]
)
case_summary = case_summary.with_columns(
(
(
pl.col(case_end_col).dt.timestamp("ns")
- pl.col(case_start_col).dt.timestamp("ns")
)
/ 1_000_000_000
).alias("@@case_throughput")
)
if compute_extra_temporal_features:
case_summary = case_summary.with_columns(
[
pl.col(case_start_col).dt.strftime("%Y").alias("@@case_start_year"),
pl.col(case_start_col).dt.strftime("%Y-%m").alias("@@case_start_ymonth"),
pl.concat_str(pl.lit("M"), pl.col(case_start_col).dt.strftime("%m")).alias(
"@@case_start_month"
),
pl.concat_str(
pl.lit("W"),
pl.col(case_start_col)
.dt.week()
.cast(pl.Utf8)
.str.pad_start(2, "0"),
).alias("@@case_start_week"),
pl.col(case_end_col).dt.strftime("%Y").alias("@@case_end_year"),
pl.col(case_end_col).dt.strftime("%Y-%m").alias("@@case_end_ymonth"),
pl.concat_str(pl.lit("M"), pl.col(case_end_col).dt.strftime("%m")).alias(
"@@case_end_month"
),
pl.concat_str(
pl.lit("W"),
pl.col(case_end_col)
.dt.week()
.cast(pl.Utf8)
.str.pad_start(2, "0"),
).alias("@@case_end_week"),
]
)
select_columns: List[pl.Expr] = [pl.col(case_id_key), pl.col("@@case_throughput")]
if compute_extra_temporal_features:
select_columns.extend(
[
pl.col("@@case_start_year"),
pl.col("@@case_start_ymonth"),
pl.col("@@case_start_month"),
pl.col("@@case_start_week"),
pl.col("@@case_end_year"),
pl.col("@@case_end_ymonth"),
pl.col("@@case_end_month"),
pl.col("@@case_end_week"),
]
)
return case_summary.select(select_columns)
[docs]
def compute_extra_columns(
dataframe: pl.LazyFrame,
parameters: Optional[Dict[Any, Any]] = None,
) -> pl.LazyFrame:
"""Enrich a Polars LazyFrame with additional case-level columns."""
if parameters is None:
parameters = {}
case_id_key = exec_utils.get_param_value(
Parameters.CASE_ID_KEY, parameters, constants.CASE_CONCEPT_NAME
)
start_timestamp_key = exec_utils.get_param_value(
Parameters.START_TIMESTAMP_KEY,
parameters,
xes_constants.DEFAULT_TIMESTAMP_KEY,
)
timestamp_key = exec_utils.get_param_value(
Parameters.TIMESTAMP_KEY,
parameters,
xes_constants.DEFAULT_TIMESTAMP_KEY,
)
compute_extra_temporal_features = exec_utils.get_param_value(
Parameters.COMPUTE_EXTRA_TEMPORAL_FEATURES, parameters, True
)
# Drop any previously-added enrichment columns to avoid duplicated columns
# (including join-suffix leftovers from older versions).
all_enrichment_columns = [
"@@count",
"@@case_throughput",
"@@case_start_year",
"@@case_start_ymonth",
"@@case_start_month",
"@@case_start_week",
"@@case_end_year",
"@@case_end_ymonth",
"@@case_end_month",
"@@case_end_week",
]
drop_candidates = list(all_enrichment_columns) + [
f"{c}_right" for c in all_enrichment_columns
]
dataframe = _drop_if_present(dataframe, drop_candidates)
df = dataframe.with_columns(pl.lit(1).alias("@@count"))
case_features = _prepare_case_features(
df,
case_id_key,
start_timestamp_key,
timestamp_key,
compute_extra_temporal_features,
)
enriched = df.join(case_features, on=case_id_key, how="left", coalesce=True)
return enriched
__all__ = ["Parameters", "compute_extra_columns"]
