main.py

import os
os.environ['TF_ENABLE_ONEDNN_OPTS'] = '0'
import silence_tensorflow.auto
from tensorflow_utilities.tensor_model import TensorModel
from binance.client import Client
from binanceAPI.position_utilities import enter_long, enter_short
from config import api_key, secret_key
from indicators.fetch_all_indicators import fetch_all_indicators
from data.io_utilities import print_with_color, calculateWR, print_position_message
from time import sleep
from data.data_functions import save_position, save_result
import copy

# Binance API initialization
client = Client(api_key, secret_key)

# Global Variables
csv_path_position = "./data/sagebot_dataset.csv"
csv_path_result = "./data/sagebot_result.csv"
data_objects = [None, None]
on_long = False
on_short = False
tp_price = 0
sl_price = 0
prediction = None
tp_count = 0
sl_count = 0

# Global Functions
def close_position(isTP):
    global on_long
    global on_short
    global tp_count
    global sl_count
    global data_objects
    global prediction
    global csv_path_position
    global csv_path_result

    state = "" 
    
    if (on_long and isTP) or (on_short and (not isTP)): 
        state = "LONG"
    elif (on_long and (not isTP)) or (on_short and isTP):
        state = "SHORT"

    save_position(csv_path_position, state, data_objects[0])
    save_result(csv_path_result, data_objects[0].date, state, prediction)

    on_long = False
    on_short = False

    if isTP:
        tp_count = tp_count + 1
        print_with_color("green", "Position closed with TP")
        print_with_color("yellow", "TP: " + str(tp_count) + " SL: " + 
              str(sl_count) + " Win-Rate: " + calculateWR(tp_count, sl_count))
    else:
        sl_count = sl_count + 1
        print_with_color("red", "Position closed with SL")
        print_with_color("yellow", "TP: " + str(tp_count) + " SL: " + 
            str(sl_count) + " Win-Rate: " + calculateWR(tp_count, sl_count))
        
print_with_color("cyan", "SageBot is running...")

while True:
    try:
        sleep(10)
        data_objects[1] = fetch_all_indicators(client)

        if not (on_long or on_short):
            print()
            data_objects[0] = copy.deepcopy(data_objects[1])
            accuracy, prediction = TensorModel(csv_path_position).process_model(data_objects[0])
            if prediction == "LONG":
                tp_price, sl_price = enter_long(client)
                on_long = True
            elif prediction == "SHORT":
                tp_price, sl_price = enter_short(client)
                on_short = True
            
            print_with_color("yellow", "Entered " + prediction + " Current: " + 
                             str(round(data_objects[0].price, 2)) + " TP_PRICE: " + str(round(tp_price, 2)) + 
                             " SL_PRICE: " + str(round(sl_price, 2)) + " ACCURACY: " + 
                             (str(round(accuracy, 2)) if accuracy is not None else "None"))
            print_position_message(data_objects[0], prediction)
        else:
            if (on_long and  data_objects[1].price > tp_price) or \
                  (on_short and data_objects[1].price < tp_price):
                close_position(True)
            elif (on_long and data_objects[1].price < sl_price) or \
                  (on_short and data_objects[1].price > sl_price):
                close_position(False)

    except Exception as e:
        error_message = str(e)
        print_with_color("yellow", error_message)