Merge pull request #10 from napo/main

merge pull requests

datavolley/dataaugmentation.py

@@ -0,0 +1,96 @@
+"""module to enrich data"""
+import pandas as pd
+import numpy as np
+
+# Some examples of adding extra columns to our play-by-play data in order 
+#to support particular analyses, or to make other data-wrangling tasks easier.
+# Aim:
+# Identify the player who made the set associated with each attack 
+# (nothing that some files might not have the setting action coded
+# for all attacks, or even coded at all).
+
+def setterofattackplays(plays):
+    """
+    This function identifies the setter 
+    for each attack play in a volleyball match.
+
+    It adds a new column 'set_player_name' to the DataFrame, 
+    which contains the name of the player
+    who performed the set immediately before the attack, 
+    if the set and attack were made by the
+    same team. If these conditions are not met, the value is set to NaN.
+
+    Args:
+    plays (DataFrame): 
+    A pandas DataFrame containing volleyball play data, with at least the columns
+        'skill', 'team', and 'player_name'.
+
+    Returns:
+    DataFrame: 
+    A filtered DataFrame containing only the first 5 rows of attack plays, with columns
+        'team', 'player_name', 'skill', 'evaluation_code', and 'set_player_name'.
+    """
+    plays['set_player_name'] = np.where((plays['skill'] == 'Attack') &
+                                        (plays['skill'].shift(1) == 'Set') &
+                                        (plays['team'] == plays['team'].shift(1)),
+                                        plays['player_name'].shift(1), np.nan)
+    filtered_plays = \
+        plays[plays['skill'] == 'Attack']\
+            [['team', 'player_name',\
+               'skill', 'evaluation_code', 'set_player_name']].head(5)
+    return filtered_plays
+
+def setplayerid(plays):
+    """
+    Identifies the player_id of the setter on court for each data row.
+
+    This function adds a new column 'setter_id' to the DataFrame, which contains the player_id
+    of the setter based on the 'setter_position' value. If 'setter_position' is 0 or if there's 
+    an issue accessing the player_id, the value is set to None.
+
+    Args:
+    plays (DataFrame): A pandas DataFrame containing volleyball play data with columns 
+                       'setter_position' and 'home_pX' where X is a position number.
+
+    Returns:
+    DataFrame: The modified DataFrame with an additional column 'setter_id'.
+    """
+    def assign_setter_id(row):
+        if row['setter_position'] == 0:
+            return None
+        try:
+            return row['home_p' + str(row['setter_position'])]
+        except KeyError:
+            return None
+
+    plays['setter_id'] = plays.apply(assign_setter_id, axis=1)
+    return plays
+
+
+def receptionquality(plays):
+    """
+    Adds a column that indicates the reception quality associated with each rally.
+
+    This function identifies the reception quality for each rally based on the 'skill' column.
+    It adds a new column 'reception_quality' to the DataFrame, which contains the evaluation of
+    the reception if there is exactly one reception per rally. If there are multiple receptions
+    or no receptions in a rally, the value is set to None.
+
+    Args:
+    plays (DataFrame): A pandas DataFrame containing volleyball play data with columns 'set_number',
+                       'rally_number', 'skill', and 'evaluation'.
+
+    Returns:
+    DataFrame: The modified DataFrame with an additional column 'reception_quality'.
+    """
+    reception_rows = plays[plays['skill'] == 'Reception']
+    grouped_reception = reception_rows.groupby(['set_number', 'rally_number'])
+
+    def calculate_reception_quality(group):
+        if len(group) == 1:
+            return group['evaluation'].iloc[0]
+        return None
+
+    rq = grouped_reception.apply(calculate_reception_quality).reset_index(name='reception_quality')
+    plays = pd.merge(plays, rq, on=['set_number', 'rally_number'], how='left')
+    return plays