The basic architecture of the application is built.

.env.example

@@ -0,0 +1,5 @@
+WEIGHTS_USE=True
+
+KNN_WEIGHT=0.8
+LOGISTICS_REGRESSION_WEIGHT=0.7
+XGBOOST_WEIGHT=1

disk_destiny/app.py

@@ -0,0 +1,87 @@
+import os
+import inquirer
+import numpy as np
+from dotenv import load_dotenv
+from typing import Dict, List
+
+import models
+from models.base import BaseModel
+from utils.other import print_pure_banner
+
+
+class App:
+	allowed_models: Dict[str, BaseModel]
+
+	def __init__(self, weights: Dict[str, float], weights_use: bool = True) -> None:
+		self.allowed_models = {}
+
+		for model in BaseModel.__subclasses__():
+			weight =  weights.get(model.name.upper(), 1)
+
+			try:
+				weight = float(weight)
+			except ValueError:
+				raise ValueError(f"Вес для модели {model.name} должен быть числом")
+
+			if weight < 0 or weight > 1:
+				raise ValueError(f"Вес для модели {model.name} должен быть от 0 до 1")
+
+			self.allowed_models[model.name] = model(weight=weight)
+
+	def run(self) -> None:
+		print_pure_banner()
+		self.ask_questions()
+
+	def ask_questions(self) -> None:
+		main_menu_answer = inquirer.prompt([
+			inquirer.List('answer',
+				message="Что будем делать?",
+				choices=['Дообучение модели', 'Получить предсказание', 'Выход'],
+			)
+		])['answer']
+
+		print_pure_banner()
+
+		match main_menu_answer:
+			case 'Дообучение модели':
+				model = inquirer.prompt([
+					inquirer.List('answer',
+						message="Выберите модель для дообучения",
+						choices=self.allowed_models,
+					)
+				])['answer']
+				self.allowed_models[model].train()
+
+			case 'Получить предсказание':
+				models = inquirer.prompt([
+					inquirer.Checkbox('answer',
+						message="Выберите модели для предсказания.",
+						choices=self.allowed_models,
+					)
+				])['answer']
+				print(models)
+				self.predict(list_models=models)
+
+			case 'Выход':
+				exit()
+
+	def predict(self, list_models: List[BaseModel], use_weights: bool = True) -> float:
+		results = {i.weight: i.predict() for i in list_models}
+
+		if use_weights:
+			return np.average(results.values(), axis=0, weights=results.keys())
+		else:
+			return np.mean(results.values(), axis=0)
+
+
+if __name__ == "__main__":
+	load_dotenv()
+	weights_use = os.environ.get('WEIGHTS_USE', 'true').lower() == 'true'
+
+	weights = {
+		key.replace("_WEIGHT", ""): value 
+		for key, value in os.environ.items() if key.endswith('_WEIGHT')
+	}
+
+	app = App(weights=weights, weights_use=weights_use)
+	app.run()

disk_destiny/models/KNN.py

@@ -0,0 +1,11 @@
+from .base import BaseModel
+
+
+class KNN(BaseModel):
+	name = 'KNN'
+
+	def train(self) -> None:
+		pass
+
+	def predict(self) -> float:
+		pass

disk_destiny/models/XGBoost.py

@@ -0,0 +1,11 @@
+from .base import BaseModel
+
+
+class XGBoost(BaseModel):
+	name = 'XGBoost'
+
+	def train(self) -> None:
+		pass
+
+	def predict(self) -> float:
+		pass

disk_destiny/models/__init__.py

@@ -0,0 +1,7 @@
+from .base import BaseModel
+from .KNN import KNN
+from .logistics_regression import LogisticsRegression
+from .XGBoost import XGBoost
+
+
+__all__ = ['BaseModel', 'KNN', 'LogisticsRegression', 'XGBoost']

disk_destiny/models/base.py

@@ -0,0 +1,15 @@
+from abc import ABC, abstractmethod
+
+
+class BaseModel(ABC):
+	name: str
+	weight: float
+
+	def __init__(self, weight: float = 1) -> None:
+		self.weight = weight
+
+	@abstractmethod
+	def train(self) -> None: ...
+
+	@abstractmethod
+	def predict(self) -> float: ...

disk_destiny/models/logistics_regression.py

@@ -0,0 +1,11 @@
+from .base import BaseModel
+
+
+class LogisticsRegression(BaseModel):
+	name = 'Logistics_Regression'
+
+	def train(self) -> None:
+		pass
+
+	def predict(self) -> float:
+		pass

disk_destiny/utils/banner.py

@@ -0,0 +1,8 @@
+banner = """
+╭━━━┳━━┳━━━┳╮╭━╮╭━━━┳━━━┳━━━┳━━━━┳━━┳━╮╱╭┳╮╱╱╭╮
+╰╮╭╮┣┫┣┫╭━╮┃┃┃╭╯╰╮╭╮┃╭━━┫╭━╮┃╭╮╭╮┣┫┣┫┃╰╮┃┃╰╮╭╯┃
+╱┃┃┃┃┃┃┃╰━━┫╰╯╯╱╱┃┃┃┃╰━━┫╰━━╋╯┃┃╰╯┃┃┃╭╮╰╯┣╮╰╯╭╯
+╱┃┃┃┃┃┃╰━━╮┃╭╮┃╱╱┃┃┃┃╭━━┻━━╮┃╱┃┃╱╱┃┃┃┃╰╮┃┃╰╮╭╯
+╭╯╰╯┣┫┣┫╰━╯┃┃┃╰╮╭╯╰╯┃╰━━┫╰━╯┃╱┃┃╱╭┫┣┫┃╱┃┃┃╱┃┃
+╰━━━┻━━┻━━━┻╯╰━╯╰━━━┻━━━┻━━━╯╱╰╯╱╰━━┻╯╱╰━╯╱╰╯
+"""

disk_destiny/utils/other.py

@@ -0,0 +1,15 @@
+import inspect
+import subprocess
+from .banner import banner
+
+
+def clear_screen() -> None:
+    subprocess.run('clear')
+
+def print_banner() -> None:
+    print(banner)
+
+def print_pure_banner() -> None:
+    clear_screen()
+    print_banner()
+