Datasets:
File size: 7,861 Bytes
0f99721 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 | """Pipeline helpers pour feature engineering - ISO 5055.
Ce module contient les fonctions d'orchestration du pipeline de features:
- extract_all_features: Extraction de toutes les features depuis l'historique
- merge_all_features: Merge de toutes les features sur le DataFrame cible
Ces fonctions sont extraites de feature_engineering.py pour respecter
la limite de 300 lignes (ISO 5055).
Conformite ISO/IEC:
- 5055: Code maintainable (<300 lignes, SRP)
- 5259: Qualite donnees ML
"""
from __future__ import annotations
import pandas as pd
from scripts.features.advanced import (
calculate_elo_trajectory,
calculate_head_to_head,
calculate_pressure_performance,
)
from scripts.features.club_behavior import extract_club_behavior
from scripts.features.composition import extract_composition_strategy
from scripts.features.ffe_features import extract_ffe_regulatory_features
from scripts.features.merge_helpers import (
merge_club_reliability,
merge_h2h_features,
merge_noyau_features,
merge_player_features,
merge_team_enjeu,
)
from scripts.features.noyau import extract_noyau_features
from scripts.features.performance import (
calculate_board_position,
calculate_color_performance,
calculate_recent_form,
)
from scripts.features.pipeline_extended import (
extract_ali_features,
merge_ali_features,
)
from scripts.features.reliability import (
extract_club_reliability,
extract_player_reliability,
)
from scripts.features.standings import calculate_standings, extract_team_enjeu_features
def extract_all_features(
df_history: pd.DataFrame,
df_history_played: pd.DataFrame,
include_advanced: bool,
) -> dict[str, pd.DataFrame]:
"""Extrait toutes les features de l'historique.
Args:
----
df_history: DataFrame historique complet
df_history_played: DataFrame parties jouees uniquement (sans forfaits)
include_advanced: Inclure features avancees (H2H, fatigue, etc.)
Returns:
-------
Dict[nom_feature, DataFrame] avec toutes les features calculees
"""
# Calcul classement pour zones enjeu (ISO 5259 - position reelle)
standings = calculate_standings(df_history_played)
features = {
"club_reliability": extract_club_reliability(df_history),
"player_reliability": extract_player_reliability(df_history),
"recent_form": calculate_recent_form(df_history_played),
"board_position": calculate_board_position(df_history_played),
"color_perf": calculate_color_performance(df_history_played),
"ffe_regulatory": extract_ffe_regulatory_features(df_history_played),
"team_enjeu": extract_team_enjeu_features(df_history_played, standings),
"club_behavior": extract_club_behavior(df_history),
"noyau": extract_noyau_features(df_history_played),
}
# ALI features (presence + patterns + absence)
ali_features = extract_ali_features(df_history_played)
features.update(ali_features)
if include_advanced:
# Composition strategy (A02 Art. 3.6.e)
compo_raw = extract_composition_strategy(df_history_played)
compo_agg = _aggregate_composition(compo_raw)
features.update(
{
"h2h": calculate_head_to_head(df_history_played),
"pressure": calculate_pressure_performance(df_history_played),
"trajectory": calculate_elo_trajectory(df_history_played),
"composition": compo_agg,
}
)
return features
def _aggregate_composition(compo_raw: pd.DataFrame) -> pd.DataFrame:
"""Agrège les features composition par joueur (moyenne historique)."""
if compo_raw.empty:
return pd.DataFrame(
columns=["joueur_nom", "decalage_position", "joueur_decale_haut", "joueur_decale_bas"]
)
return (
compo_raw.groupby("nom")
.agg(
decalage_position=("decalage_position", "mean"),
joueur_decale_haut=("joueur_decale_haut", "mean"),
joueur_decale_bas=("joueur_decale_bas", "mean"),
)
.reset_index()
.rename(columns={"nom": "joueur_nom"})
)
def merge_all_features(
result: pd.DataFrame,
features: dict[str, pd.DataFrame],
include_advanced: bool,
) -> pd.DataFrame:
"""Merge toutes les features sur le DataFrame cible.
Args:
----
result: DataFrame cible (copie du split)
features: Dict des features extraites par extract_all_features
include_advanced: Inclure features avancees
Returns:
-------
DataFrame avec toutes les features mergees
"""
# Club reliability
result = merge_club_reliability(result, features["club_reliability"])
# Player-based features
result = merge_player_features(
result, features["player_reliability"], ["taux_presence", "joueur_fantome"]
)
result = merge_player_features(
result, features["recent_form"], ["forme_recente", "forme_tendance"]
)
result = merge_player_features(
result, features["board_position"], ["echiquier_moyen", "echiquier_std"]
)
result = merge_player_features(
result,
features["color_perf"],
["score_blancs", "score_noirs", "couleur_preferee", "data_quality"],
)
# FFE regulatory features
result = merge_player_features(
result,
features["ffe_regulatory"],
["nb_equipes", "niveau_max", "niveau_min", "multi_equipe"],
prefix="ffe_",
)
# Team enjeu
result = merge_team_enjeu(result, features["team_enjeu"])
# Club behavior (merge by equipe_dom AND equipe_ext)
result = _merge_club_behavior(result, features.get("club_behavior", pd.DataFrame()))
# Noyau features (joueur x equipe x ronde)
result = merge_noyau_features(result, features.get("noyau", pd.DataFrame()))
# ALI features (presence + patterns + absence per player)
result = merge_ali_features(result, features)
# Advanced features
if include_advanced:
result = _merge_advanced_features(result, features)
return result
def _merge_club_behavior(
result: pd.DataFrame,
club_beh: pd.DataFrame,
) -> pd.DataFrame:
"""Merge club behavior pour equipe_dom ET equipe_ext.
Colonnes dom: nb_joueurs_utilises_dom, rotation_effectif_dom, etc.
Colonnes ext: nb_joueurs_utilises_ext, rotation_effectif_ext, etc.
"""
if club_beh.empty or "equipe" not in club_beh.columns:
return result
beh_cols = [c for c in club_beh.columns if c not in ("equipe", "saison")]
for suffix, equipe_col in [("dom", "equipe_dom"), ("ext", "equipe_ext")]:
if equipe_col not in result.columns:
continue
rename_map = {c: f"{c}_{suffix}" for c in beh_cols}
merge_df = club_beh.rename(columns={"equipe": equipe_col} | rename_map)
result = result.merge(
merge_df[[equipe_col, "saison"] + list(rename_map.values())],
on=[equipe_col, "saison"],
how="left",
)
return result
def _merge_advanced_features(
result: pd.DataFrame,
features: dict[str, pd.DataFrame],
) -> pd.DataFrame:
"""Merge les features avancees (H2H, pressure, trajectory, composition)."""
result = merge_player_features(
result,
features.get("trajectory", pd.DataFrame()),
["elo_trajectory", "momentum"],
)
result = merge_player_features(
result,
features.get("pressure", pd.DataFrame()),
["clutch_factor", "pressure_type"],
)
result = merge_h2h_features(result, features.get("h2h", pd.DataFrame()))
result = merge_player_features(
result,
features.get("composition", pd.DataFrame()),
["decalage_position", "joueur_decale_haut", "joueur_decale_bas"],
)
return result
|