db_logic.py

# # from flask import Flask, render_template, request, redirect, url_for, abort
# from sklearn.neighbors import KNeighborsClassifier
# from sklearn.preprocessing import StandardScaler
# from sqlalchemy.exc import SQLAlchemyError
#
# # from main import Cats, db
#
#
# # def display_all(db_model: Cats):
# #     return db_model.query.all()
#
# def display_all(Cats):
#     return Cats.query.all()
#
#
# def add(weight: float, tail_length: float, breed: int, Cats, db):
#     cat = Cats(weight=weight, tail_length=tail_length, breed=breed)
#
#     try:
#         db.session.add(cat)
#         db.session.commit()
#         # return 200  # redirect(url_for("home")) or {"id": new_cat.id}
#         return Cats.query.get(cat.id)
#     except SQLAlchemyError:
#         db.session.rollback()
#         # return 500  # abort(500, "Error while adding new record") or {"error": "Error while adding new record"}, 500
#         return None
#
#
# def delete(cat_id: int, db):
#     try:
#         db.session.delete(cat_id)
#         db.session.commit()
#         # return 200  # redirect(url_for("home")) or {"id": record_id}
#         return cat_id
#     except SQLAlchemyError:
#         db.session.rollback()
#         # return 500  # abort(500, "Error while deleting record") or {"error": "Error while deleting record"}
#         return None
#
#
# def predict(weight: float, tail_length: float, Cats):
#     k = 3
#     if Cats.query.count() < k:
#         return None
#         # abort(500, f"Not enough records in Cats database, must be at least: {k}") or
#         # {"error": f"Not enough records in Cats database, must be at least: {k}"}, 500
#
#     new_features = [[weight, tail_length]]
#
#     cats_db = Cats.query.all()
#     X = [[cat.weight, cat.tail_length] for cat in cats_db]
#     y = [cat.breed for cat in cats_db]
#
#     scaler = StandardScaler()
#     X_scaled = scaler.fit_transform(X)
#
#     new_features_scaled = scaler.transform(new_features)
#
#     neigh = KNeighborsClassifier(n_neighbors=k)
#     neigh.fit(X_scaled, y)
#     prediction = neigh.predict(new_features_scaled)[0]  # czy rzucać to na int?
#
#     return prediction