#!/usr/bin/env python3
"""
Detect bull/bear/range regimes from historical CSV data (SMA50/SMA200 crossovers)
Usage: python3 detect_regime.py <PAIR> <CSV_FILE>
CSV must contain at least columns: timestamp, close
"""

# python3 user_data/strategies/tools/detect_regime.py BTC /home/jerome/Perso/freqtradeDocker/user_data/data/binance/BTC_USDT-1d.feather

import sys
import pandas as pd
import ta
import talib.abstract as talib

if len(sys.argv) < 3:
    print("Usage: detect_regime.py <PAIR> <FEATHER_FILE>")
    sys.exit(1)

pair = sys.argv[1]
file = sys.argv[2]

# lecture du fichier feather
df = pd.read_feather(file)
# ne garder que timestamp et close pour detect_regime.py
df[['date', 'close']].rename(columns={'date':'timestamp'}).to_csv(
    f"user_data/data/{pair}-usdt-1d.csv", index=False
)

df = pd.read_csv(f"user_data/data/{pair}-usdt-1d.csv")

if 'close' not in df.columns or 'timestamp' not in df.columns:
    print("CSV must contain 'timestamp' and 'close' columns")
    sys.exit(1)

# --- paramètres ---
SMOOTH_WIN = 10   # EMA pour lisser la pente
NUM_TOP = 1      # nombre de segments à garder par classe

# --- charger les données ---
df['timestamp'] = pd.to_datetime(df['timestamp'], errors='coerce')

# filtrer uniquement 2024 et 2025
df = df[df['timestamp'].dt.year.isin([2024, 2025])]

# --- calcul SMA14 ---
df['sma'] = talib.SMA(df, timeperiod=20) #ta.trend.sma_indicator(df['close'], 14)

# --- pente brute ---
df['slope'] = df['sma'].diff()

# --- lissage EMA ---
df['slope_smooth'] = df['slope'].ewm(span=SMOOTH_WIN, adjust=False).mean()

#df["slope_smooth"] = savgol_filter(df["slope_smooth"], window_length=21, polyorder=3)

# --- normalisation relative ---
df['slope_norm'] = df['slope_smooth'] / df['close']
# df['slope_norm'].fillna(0, inplace=True)
df['slope_norm'] = df['slope_norm'].fillna(0)


# --- classification dynamique via quantiles ---

q = df['slope_norm'].quantile([0.125, 0.375, 0.625, 0.875]).values
q1, q2, q3, q4 = q

def classify(v):
    if v <= q1:
        return 'BR2'
    elif v <= q2:
        return 'BR1'
    elif v <= q3:
        return 'RG'
    elif v <= q4:
        return 'BU1'
    else:
        return 'BU2'

# q = df['slope_norm'].quantile([0.7, 0.21, 0.35, 0.65, 0.79, 0.93]).values
# q1, q2, q3, q4, q5, q6 = q
#
# def classify(v):
#     if v <= q1:
#         return 'BR3'
#     elif v <= q2:
#         return 'BR2'
#     elif v <= q3:
#         return 'BR1'
#     elif v <= q4:
#         return 'RG'
#     elif v <= q5:
#         return 'BU1'
#     elif v <= q6:
#         return 'BU2'
#     else:
#         return 'BU3'

df['trend_class'] = df['slope_norm'].apply(classify)

# --- boucle pour détecter les segments ---
segments = []
trend = None
start_idx = None

for i in range(len(df)):
    new_trend = df['trend_class'].iloc[i]

    if trend is None:
        trend = new_trend
        start_idx = i
    elif new_trend != trend:
        start_ts = df['timestamp'].iloc[start_idx]
        end_ts = df['timestamp'].iloc[i]
        segments.append((trend, start_ts, end_ts))
        trend = new_trend
        start_idx = i

# fermer le dernier segment
if trend is not None and start_idx is not None:
    start_ts = df['timestamp'].iloc[start_idx]
    end_ts = df['timestamp'].iloc[len(df)-1]
    segments.append((trend, start_ts, end_ts))

# --- extraire les 5 plus longs segments par classe ---
top_segments_by_class = {}
for cls in ['BR2','BR1','RG','BU1','BU2']:
    cls_segments = [(t,s,e) for t,s,e in segments if t==cls]
    # calcul de la durée
    cls_segments = [(t,s,e,(e-s).total_seconds()) for t,s,e in cls_segments]
    cls_segments.sort(key=lambda x: x[3], reverse=True)
    top_segments_by_class[cls] = [(t,s,e) for t,s,e,d in cls_segments[:NUM_TOP]]

# --- affichage ---
for cls, segs in top_segments_by_class.items():
    print(f"--- {cls} ---")
    for t,s,e in segs:
        print(f"{t} {s} {e}")