pm4py.objects.log.util.dataframe_utils module#

class pm4py.objects.log.util.dataframe_utils.Parameters(*values)[source]#

Bases: Enum

PARTITION_COLUMN = 'partition_column'#
CASE_ID_KEY = 'pm4py:param:case_id_key'#
CASE_PREFIX = 'case:'#
CASE_ATTRIBUTES = 'case_attributes'#
MANDATORY_ATTRIBUTES = 'mandatory_attributes'#
MAX_NO_CASES = 'max_no_cases'#
MIN_DIFFERENT_OCC_STR_ATTR = 5#
MAX_DIFFERENT_OCC_STR_ATTR = 50#
TIMESTAMP_KEY = 'pm4py:param:timestamp_key'#
ACTIVITY_KEY = 'pm4py:param:activity_key'#
PARAM_ARTIFICIAL_START_ACTIVITY = 'pm4py:param:art_start_act'#
PARAM_ARTIFICIAL_END_ACTIVITY = 'pm4py:param:art_end_act'#
INDEX_KEY = 'index_key'#
CASE_INDEX_KEY = 'case_index_key'#
USE_EXTREMES_TIMESTAMP = 'use_extremes_timestamp'#
ADD_CASE_IDENTIFIER_COLUMN = 'add_case_identifier_column'#
DETERMINISTIC = 'deterministic'#
pm4py.objects.log.util.dataframe_utils.insert_partitioning(df, num_partitions, parameters=None)[source]#

Insert the partitioning in the specified dataframe

Parameters:

  • df – Dataframe

  • num_partitions – Number of partitions

  • parameters – Parameters of the algorithm

Returns:

Partitioned dataframe

Return type:

df

pm4py.objects.log.util.dataframe_utils.legacy_parquet_support(df, parameters=None)[source]#

For legacy support, Parquet files columns could not contain a “:” that has been arbitrarily replaced by a replacer string. This string substitutes the replacer to the :

Parameters:

  • dataframe – Dataframe

  • parameters – Parameters of the algorithm

pm4py.objects.log.util.dataframe_utils.table_to_stream(table, parameters=None)[source]#

Converts a Pyarrow table to an event stream

Parameters:

  • table – Pyarrow table

  • parameters – Possible parameters of the algorithm

pm4py.objects.log.util.dataframe_utils.table_to_log(table, parameters=None)[source]#

Converts a Pyarrow table to an event log

Parameters:

  • table – Pyarrow table

  • parameters – Possible parameters of the algorithm

pm4py.objects.log.util.dataframe_utils.convert_timestamp_columns_in_df(df, timest_format=None, timest_columns=None)[source]#

Convert all dataframe columns in a dataframe

Parameters:

  • df – Dataframe

  • timest_format – (If provided) Format of the timestamp columns in the CSV file

  • timest_columns – Columns of the CSV that shall be converted into timestamp

Returns:

Dataframe with timestamp columns converted

Return type:

df

pm4py.objects.log.util.dataframe_utils.sample_dataframe(df, parameters=None)[source]#

Sample a dataframe on a given number of cases

Parameters:

  • df – Dataframe

  • parameters – Parameters of the algorithm, including: - Parameters.CASE_ID_KEY - Parameters.CASE_ID_TO_RETAIN

Returns:

Sampled dataframe

Return type:

sampled_df

pm4py.objects.log.util.dataframe_utils.automatic_feature_selection_df(df, parameters=None)[source]#

Performs an automatic feature selection on dataframes, keeping the features useful for ML purposes

Parameters:

  • df – Dataframe

  • parameters – Parameters of the algorithm

Returns:

Dataframe with only the features that have been selected

Return type:

featured_df

pm4py.objects.log.util.dataframe_utils.select_number_column(df: DataFrame, fea_df: DataFrame, col: str, case_id_key='case:concept:name') → DataFrame[source]#

Extract a column for the features dataframe for the given numeric attribute

Parameters:

  • df – Dataframe

  • fea_df – Feature dataframe

  • col – Numeric column

  • case_id_key – Case ID key

Returns:

Feature dataframe (desidered output)

Return type:

fea_df

pm4py.objects.log.util.dataframe_utils.select_string_column(df: DataFrame, fea_df: DataFrame, col: str, case_id_key='case:concept:name') → DataFrame[source]#

Extract N columns (for N different attribute values; hotencoding) for the features dataframe for the given string attribute

Parameters:

  • df – Dataframe

  • fea_df – Feature dataframe

  • col – String column

  • case_id_key – Case ID key

Returns:

Feature dataframe (desidered output)

Return type:

fea_df

pm4py.objects.log.util.dataframe_utils.get_features_df(df: DataFrame, list_columns: List[str], parameters: Dict[Any, Any] | None = None) → DataFrame[source]#

Given a dataframe and a list of columns, performs an automatic feature extraction

Parameters:

  • df – Dataframe

  • list_column – List of column to consider in the feature extraction

  • parameters – Parameters of the algorithm, including: - Parameters.CASE_ID_KEY: the case ID

Returns:

Feature dataframe (desidered output)

Return type:

fea_df

pm4py.objects.log.util.dataframe_utils.automatic_feature_extraction_df(df: DataFrame, parameters: Dict[Any, Any] | None = None) → DataFrame[source]#

Performs an automatic feature extraction given a dataframe

Parameters:

  • df – Dataframe

  • parameters – Parameters of the algorithm, including: - Parameters.CASE_ID_KEY: the case ID - Parameters.MIN_DIFFERENT_OCC_STR_ATTR - Parameters.MAX_DIFFERENT_OCC_STR_ATTR

Returns:

Dataframe with the features

Return type:

fea_df

pm4py.objects.log.util.dataframe_utils.insert_artificial_start_end(df0: DataFrame, parameters: Dict[Any, Any] | None = None) → DataFrame[source]#

Inserts the artificial start/end activities in a Pandas dataframe

Parameters:

  • df0 – Dataframe

  • parameters – Parameters of the algorithm, including: - Parameters.CASE_ID_KEY: the case identifier - Parameters.TIMESTAMP_KEY: the timestamp - Parameters.ACTIVITY_KEY: the activity

Returns:

Dataframe with artificial start/end activities

Return type:

enriched_df

pm4py.objects.log.util.dataframe_utils.dataframe_to_activity_case_table(df: DataFrame, parameters: Dict[Any, Any] | None = None)[source]#

Transforms a Pandas dataframe into: - an “activity” table, containing the events and their attributes - a “case” table, containing the cases and their attributes

Parameters:

  • df – Dataframe

  • parameters – Parameters of the algorithm that should be used, including: - Parameters.CASE_ID_KEY => the column to be used as case ID (shall be included both in the activity table and the case table) - Parameters.CASE_PREFIX => if a list of attributes at the case level is not provided, then all the ones of the dataframe

    starting with one of these are considered.

    • Parameters.CASE_ATTRIBUTES => the attributes of the dataframe to be used as case columns

Returns:

  • activity_table – Activity table

  • case_table – Case table