Merge branch 'PM4PY-1563' into 'integration'

PMPY-1563 Added "week of the year" option in get_events_distribution method See merge request process-mining/pm4py/pm4py-core!598
process-intelligence-solutions · Jan 25, 2022 · c7c7ed5 · c7c7ed5
2 parents 2200a0f + a6bcae9
commit c7c7ed5
Show file tree

Hide file tree

Showing 3 changed files with 13 additions and 0 deletions.
diff --git a/pm4py/statistics/attributes/log/get.py b/pm4py/statistics/attributes/log/get.py
@@ -94,6 +94,9 @@ def get_events_distribution(log: EventLog, distr_type: str = "days_month", param
     elif distr_type == "days_week":
         values = Counter(x.weekday() for x in timestamp_values)
         all_values = Counter({i: 0 for i in range(0, 7)})
+    elif distr_type == "weeks":
+        values = Counter(x.isocalendar().week for x in timestamp_values)
+        all_values = Counter({i: 0 for i in range(0, 53)})
 
     # make sure that all the possible values appear
     for v in all_values:

diff --git a/pm4py/statistics/attributes/pandas/get.py b/pm4py/statistics/attributes/pandas/get.py
@@ -54,6 +54,7 @@ def get_events_distribution(df: pd.DataFrame, distr_type: str = "days_month", pa
         - years => Gets the distribution of the events among the years of the event log
         - hours => Gets the distribution of the events among the hours of a day (from 0 to 23)
         - days_week => Gets the distribution of the events among the days of a week (from Monday to Sunday)
+        - weeks => Distribution of the events among the weeks of a year (from 0 to 52)
     parameters
         Parameters of the algorithm, including:
         - Parameters.TIMESTAMP_KEY
@@ -92,6 +93,10 @@ def get_events_distribution(df: pd.DataFrame, distr_type: str = "days_month", pa
         serie = df[timestamp_key].dt.dayofweek
         values = Counter(serie.value_counts().to_dict())
         all_values = Counter({i: 0 for i in range(0, 7)})
+    elif distr_type == "weeks":
+        serie = df[timestamp_key].dt.isocalendar().week
+        values = Counter(serie.value_counts().to_dict())
+        all_values = Counter({i: 0 for i in range(0, 53)})
 
     # make sure that all the possible values appear
     for v in all_values:

diff --git a/pm4py/vis.py b/pm4py/vis.py
@@ -556,6 +556,10 @@ def __builds_events_distribution_graph(log: Union[EventLog, pd.DataFrame], distr
         title = "Distribution of the Events over the Days of a Week";
         x_axis = "Day of the Week";
         y_axis = "Number of Events"
+    elif distr_type == "weeks":
+        title = "Distribution of the Events over the Weeks of a Year";
+        x_axis = "Week of the Year";
+        y_axis = "Number of Events"
     else:
         raise Exception("unsupported distribution specified.")
 
@@ -585,6 +589,7 @@ def view_events_distribution_graph(log: Union[EventLog, pd.DataFrame], distr_typ
         - years => Gets the distribution of the events among the years of the event log
         - hours => Gets the distribution of the events among the hours of a day (from 0 to 23)
         - days_week => Gets the distribution of the events among the days of a week (from Monday to Sunday)
+        - weeks => Gets the distribution of the events among the weeks of a year (from 0 to 52)
     format
         Format of the visualization (default: png)
     """