Source code for pm4py.statistics.rework.cases.pandas.get

from enum import Enum
from typing import Optional, Dict, Any, Union

import pandas as pd
from pm4py.util import exec_utils, constants, xes_constants




class Parameters(Enum):
    ACTIVITY_KEY = constants.PARAMETER_CONSTANT_ACTIVITY_KEY
    CASE_ID_KEY = constants.PARAMETER_CONSTANT_CASEID_KEY





def apply(
    df: pd.DataFrame,
    parameters: Optional[Dict[Union[str, Parameters], Any]] = None,
) -> Dict[str, Dict[str, int]]:
    """
    Computes for each trace of the event log how much rework occurs.
    The rework is computed as the difference between the total number of activities of a trace and the
    number of unique activities.

    Parameters
    ----------------
    df
        Pandas dataframe
    parameters
        Parameters of the algorithm, including:
        - Parameters.ACTIVITY_KEY => the activity key
        - Parameters.CASE_ID_KEY => the case identifier attribute

    Returns
    -----------------
    dict
        Dictionary associating to each case ID:
        - The number of total activities of the case (number of events)
        - The rework (difference between the total number of activities of a trace and the number of unique activities)
    """
    if parameters is None:
        parameters = {}

    activity_key = exec_utils.get_param_value(
        Parameters.ACTIVITY_KEY, parameters, xes_constants.DEFAULT_NAME_KEY
    )
    case_id_key = exec_utils.get_param_value(
        Parameters.CASE_ID_KEY, parameters, constants.CASE_CONCEPT_NAME
    )

    grouped_df = (
        df.groupby(case_id_key)[activity_key]
        .agg(["count", "nunique"])
        .reset_index()
        .to_dict("records")
    )
    rework_cases = {}
    for el in grouped_df:
        rework_cases[el["case:concept:name"]] = {
            "number_activities": el["count"],
            "rework": el["count"] - el["nunique"],
        }

    return rework_cases