# 代码文件拆分说明 我理解您的需求！您希望将每个带序号的部分都拆分成单独的Python文件。以下是完整的文件结构和每个文件的代码内容： ## 📁 项目文件结构 ``` ai_basic_project/ ├── 01_environment_check.py ├── 02_data_preparation.py ├── 03_model_training.py ├── 04_inference_engine.py ├── 05_interactive_interface.py ├── 06_model_optimization.py ├── 07_model_persistence.py └── requirements.txt ``` ## 📄 各文件详细代码 ### 文件 1: 环境检查 ```python # 01_environment_check.py import sys import subprocess def check_environment(): """检查必要的库是否已安装""" required_packages = [ 'numpy', 'pandas', 'scikit-learn', 'matplotlib', 'jupyter', 'joblib' ] missing_packages = [] for package in required_packages: try: __import__(package) print(f"✅ {package} 已安装") except ImportError: missing_packages.append(package) print(f"❌ {package} 未安装") if missing_packages: print(f"\n请安装缺失的包: pip install {' '.join(missing_packages)}") return False else: print("\n🎉 环境配置完成，可以开始AI开发！") return True if __name__ == "__main__": check_environment() ``` ### 文件 2: 数据准备 ```python # 02_data_preparation.py import pandas as pd import numpy as np from sklearn.datasets import load_iris from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler import matplotlib.pyplot as plt def load_and_explore_data(): """加载和探索鸢尾花数据集""" # 加载数据集 iris = load_iris() X = iris.data y = iris.target # 创建DataFrame iris_df = pd.DataFrame(X, columns=iris.feature_names) iris_df['target'] = y iris_df['species'] = iris_df['target'].map({0: 'setosa', 1: 'versicolor', 2: 'virginica'}) print("数据集基本信息:") print(f"样本数量: {len(iris_df)}") print(f"特征数量: {X.shape[1]}") print(f"类别分布:\n{iris_df['species'].value_counts()}") return iris, X, y, iris_df def visualize_data(iris_df): """数据可视化""" plt.figure(figsize=(12, 4)) plt.subplot(1, 2, 1) for species in iris_df['species'].unique(): species_data = iris_df[iris_df['species'] == species] plt.scatter(species_data['sepal length (cm)'], species_data['sepal width (cm)'], label=species, alpha=0.7) plt.xlabel('花萼长度 (cm)') plt.ylabel('花萼宽度 (cm)') plt.legend() plt.title('花萼尺寸分布') plt.subplot(1, 2, 2) for species in iris_df['species'].unique(): species_data = iris_df[iris_df['species'] == species] plt.scatter(species_data['petal length (cm)'], species_data['petal width (cm)'], label=species, alpha=0.7) plt.xlabel('花瓣长度 (cm)') plt.ylabel('花瓣宽度 (cm)') plt.legend() plt.title('花瓣尺寸分布') plt.tight_layout() plt.savefig('data_visualization.png') plt.show() def prepare_training_data(X, y): """准备训练数据""" # 数据标准化 scaler = StandardScaler() X_scaled = scaler.fit_transform(X) # 划分训练测试集 X_train, X_test, y_train, y_test = train_test_split( X_scaled, y, test_size=0.2, random_state=42, stratify=y ) print(f"训练集大小: {X_train.shape[0]}") print(f"测试集大小: {X_test.shape[0]}") return X_train, X_test, y_train, y_test, scaler if __name__ == "__main__": iris, X, y, iris_df = load_and_explore_data() visualize_data(iris_df) X_train, X_test, y_train, y_test, scaler = prepare_training_data(X, y) ``` ### 文件 3: 模型训练 ```python # 03_model_training.py import pandas as pd import numpy as np from sklearn.datasets import load_iris from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler from sklearn.ensemble import RandomForestClassifier from sklearn.metrics import accuracy_score, classification_report, confusion_matrix import matplotlib.pyplot as plt import joblib def train_model(X_train, y_train): """训练随机森林模型""" model = RandomForestClassifier( n_estimators=100, random_state=42, max_depth=3 ) model.fit(X_train, y_train) print("✅ 模型训练完成！") return model def evaluate_model(model, X_test, y_test, iris): """评估模型性能""" # 预测 y_pred = model.predict(X_test) # 准确率 accuracy = accuracy_score(y_test, y_pred) print(f"🎯 模型准确率: {accuracy:.4f}") # 分类报告 print("\n详细分类报告:") print(classification_report(y_test, y_pred, target_names=iris.target_names)) # 混淆矩阵 plot_confusion_matrix(y_test, y_pred, iris) return accuracy, y_pred def plot_confusion_matrix(y_test, y_pred, iris): """绘制混淆矩阵""" cm = confusion_matrix(y_test, y_pred) plt.figure(figsize=(8, 6)) plt.imshow(cm, interpolation='nearest', cmap=plt.cm.Blues) plt.title('混淆矩阵') plt.colorbar() tick_marks = np.arange(len(iris.target_names)) plt.xticks(tick_marks, iris.target_names, rotation=45) plt.yticks(tick_marks, iris.target_names) plt.xlabel('预测标签') plt.ylabel('真实标签') # 在矩阵中显示数值 for i in range(cm.shape[0]): for j in range(cm.shape[1]): plt.text(j, i, format(cm[i, j], 'd'), horizontalalignment="center", color="white" if cm[i, j] > cm.max()/2. else "black") plt.tight_layout() plt.savefig('confusion_matrix.png') plt.show() if __name__ == "__main__": # 加载数据 iris = load_iris() X = iris.data y = iris.target # 数据预处理 scaler = StandardScaler() X_scaled = scaler.fit_transform(X) X_train, X_test, y_train, y_test = train_test_split( X_scaled, y, test_size=0.2, random_state=42, stratify=y ) # 训练模型 model = train_model(X_train, y_train) # 评估模型 accuracy, y_pred = evaluate_model(model, X_test, y_test, iris) # 保存模型和scaler joblib.dump({'model': model, 'scaler': scaler}, 'trained_model.pkl') print("✅ 模型已保存为 'trained_model.pkl'") ``` ### 文件 4: 推理引擎 ```python # 04_inference_engine.py import numpy as np import joblib class IrisClassifier: """鸢尾花分类器推理引擎""" def __init__(self, model_path=None): if model_path: self.load_model(model_path) else: self.model = None self.scaler = None self.feature_names = None self.target_names = None def load_model(self, model_path): """加载预训练模型""" model_data = joblib.load(model_path) self.model = model_data['model'] self.scaler = model_data['scaler'] # 加载鸢尾花数据集信息 from sklearn.datasets import load_iris iris = load_iris() self.feature_names = iris.feature_names self.target_names = iris.target_names print("✅ 模型加载成功！") def predict(self, features): """预测鸢尾花种类""" if self.model is None: raise ValueError("请先加载模型！") # 输入验证 if len(features) != 4: raise ValueError("需要提供4个特征值: 花萼长度、花萼宽度、花瓣长度、花瓣宽度") # 数据预处理 features_array = np.array(features).reshape(1, -1) features_scaled = self.scaler.transform(features_array) # 模型预测 prediction = self.model.predict(features_scaled)[0] probabilities = self.model.predict_proba(features_scaled)[0] # 返回结果 result = { 'predicted_class': self.target_names[prediction], 'confidence': probabilities[prediction], 'all_probabilities': { self.target_names[i]: prob for i, prob in enumerate(probabilities) } } return result def test_inference(): """测试推理功能""" classifier = IrisClassifier('trained_model.pkl') test_samples = [ [5.1, 3.5, 1.4, 0.2], # 应该是setosa [6.0, 2.7, 5.1, 1.6], # 应该是virginica [5.5, 2.4, 3.8, 1.1], # 应该是versicolor ] print("🧪 模型推理测试:") for i, sample in enumerate(test_samples, 1): result = classifier.predict(sample) print(f"\n样本 {i}: {sample}") print(f"预测种类: {result['predicted_class']}") print(f"置信度: {result['confidence']:.4f}") print("各类别概率:") for species, prob in result['all_probabilities'].items(): print(f" {species}: {prob:.4f}") return classifier if __name__ == "__main__": classifier = test_inference() ``` ### 文件 5: 交互界面 ```python # 05_interactive_interface.py from inference_engine import IrisClassifier import joblib def interactive_classifier(): """交互式鸢尾花分类器""" print("🌷 鸢尾花分类器 v1.0") print("=" * 40) print("请输入鸢尾花的4个特征值:") print("格式: 花萼长度 花萼宽度 花瓣长度 花瓣宽度") print("示例: 5.1 3.5 1.4 0.2") print("输入 'quit' 退出程序") print("=" * 40) # 加载模型 try: classifier = IrisClassifier('trained_model.pkl') except: print("❌ 无法加载模型，请先运行训练脚本") return while True: try: user_input = input("\n请输入特征值: ").strip() if user_input.lower() in ['quit', 'exit', 'q']: print("感谢使用鸢尾花分类器！") break # 解析输入 features = [float(x) for x in user_input.split()] if len(features) != 4: print("❌ 错误：请输入4个数值") continue # 进行预测 result = classifier.predict(features) # 显示结果 print(f"\n🎯 预测结果: {result['predicted_class']}") print(f"📊 置信度: {result['confidence']:.2%}") print("\n详细概率分布:") for species, prob in result['all_probabilities'].items(): bar = "█" * int(prob * 20) print(f" {species:12}: {prob:6.2%} {bar}") except ValueError: print("❌ 错误：请输入有效的数值") except Exception as e: print(f"❌ 发生错误: {e}") if __name__ == "__main__": interactive_classifier() ``` ### 文件 6: 模型优化 ```python # 06_model_optimization.py from sklearn.model_selection import GridSearchCV from sklearn.ensemble import RandomForestClassifier from sklearn.datasets import load_iris from sklearn.model_selection import train_test_split from sklearn.preprocessing import StandardScaler import joblib def optimize_model(X_train, y_train): """使用网格搜索优化模型参数""" param_grid = { 'n_estimators': [50, 100, 200], 'max_depth': [3, 5, 7, None], 'min_samples_split': [2, 5, 10] } grid_search = GridSearchCV( RandomForestClassifier(random_state=42), param_grid, cv=5, scoring='accuracy', n_jobs=-1 ) print("🔍 开始网格搜索优化...") grid_search.fit(X_train, y_train) print("🎯 最佳参数:", grid_search.best_params_) print("🏆 最佳准确率:", grid_search.best_score_) return grid_search.best_estimator_ if __name__ == "__main__": # 加载数据 iris = load_iris() X = iris.data y = iris.target # 数据预处理 scaler = StandardScaler() X_scaled = scaler.fit_transform(X) X_train, X_test, y_train, y_test = train_test_split( X_scaled, y, test_size=0.2, random_state=42, stratify=y ) # 优化模型 best_model = optimize_model(X_train, y_train) # 保存优化后的模型 model_data = { 'model': best_model, 'scaler': scaler, 'optimized': True } joblib.dump(model_data, 'optimized_model.pkl') print("✅ 优化后的模型已保存为 'optimized_model.pkl'") ``` ### 文件 7: 模型持久化 ```python # 07_model_persistence.py import joblib import datetime from sklearn.datasets import load_iris def save_model_with_metadata(model, scaler, accuracy, filename=None): """保存模型和元数据""" if filename is None: timestamp = datetime.datetime.now().strftime("%Y%m%d_%H%M%S") filename = f"iris_classifier_{timestamp}.pkl" # 加载数据集信息 iris = load_iris() model_data = { 'model': model, 'scaler': scaler, 'metadata': { 'feature_names': iris.feature_names.tolist(), 'target_names': iris.target_names.tolist(), 'accuracy': accuracy, 'dataset': 'iris', 'created_at': datetime.datetime.now().isoformat(), 'version': '1.0' } } joblib.dump(model_data, filename) print(f"✅ 模型已保存: {filename}") return filename def load_model_with_metadata(filename): """加载模型和元数据""" model_data = joblib.load(filename) metadata = model_data['metadata'] print(f"✅ 模型加载成功") print(f"📅 创建时间: {metadata['created_at']}") print(f"🎯 准确率: {metadata['accuracy']:.4f}") print(f"🔢 特征: {metadata['feature_names']}") print(f"🏷️ 类别: {metadata['target_names']}") return model_data def list_saved_models(): """列出所有保存的模型""" import glob model_files = glob.glob("iris_classifier_*.pkl") if model_files: print("📁 已保存的模型文件:") for file in model_files: print(f" - {file}") else: print("📁 没有找到保存的模型文件") if __name__ == "__main__": list_saved_models() # 示例：加载并显示模型信息（如果有的话） model_files = glob.glob("iris_classifier_*.pkl") if model_files: load_model_with_metadata(model_files[0]) ``` ### 依赖文件 ```txt # requirements.txt numpy>=1.21.0 pandas>=1.3.0 scikit-learn>=1.0.0 matplotlib>=3.5.0 jupyter>=1.0.0 joblib>=1.1.0 ``` ## 🚀 使用说明 1. **环境准备**: 运行 `01_environment_check.py` 2. **数据探索**: 运行 `02_data_preparation.py` 3. **模型训练**: 运行 `03_model_training.py` 4. **模型测试**: 运行 `04_inference_engine.py` 5. **交互使用**: 运行 `05_interactive_interface.py` 6. **模型优化**: 运行 `06_model_optimization.py`（可选） 7. **模型管理**: 运行 `07_model_persistence.py` 每个文件都可以独立运行，也相互关联，构成了完整的AI项目开发流程！