aigroup-econ-mcp 1.3.3__py3-none-any.whl → 2.0.1__py3-none-any.whl

econometrics/tests/specific_data_modeling_tests/test_garch.py

@@ -0,0 +1,118 @@
+"""
+GARCH模型测试脚本
+"""
+
+import sys
+import os
+import numpy as np
+
+# 添加项目根目录到路径
+sys.path.append(os.path.join(os.path.dirname(__file__), '..', '..', '..'))
+
+from econometrics.specific_data_modeling.time_series_panel_data.garch_model import garch_model, GARCHResult
+
+
+def test_garch_basic():
+    """测试基本GARCH功能"""
+    print("测试基本GARCH功能...")
+    
+    # 生成GARCH(1,1)过程数据
+    np.random.seed(42)
+    n = 500
+    
+    # GARCH(1,1)参数
+    omega = 0.1
+    alpha = 0.3
+    beta = 0.6
+    
+    # 确保参数满足平稳性条件
+    assert alpha + beta < 1, "GARCH过程需要满足alpha + beta < 1"
+    
+    # 生成GARCH过程
+    data = [0]  # 初始值
+    h = [0.5]   # 条件方差初始值
+    
+    for i in range(1, n):
+        # 生成条件方差
+        h_t = omega + alpha * (data[i-1]**2) + beta * h[i-1]
+        h.append(h_t)
+        
+        # 生成收益率
+        epsilon_t = np.random.randn() * np.sqrt(h_t)
+        data.append(epsilon_t)
+    
+    # 执行GARCH(1,1)模型
+    result = garch_model(data, order=(1, 1))
+    
+    # 验证结果类型
+    assert isinstance(result, GARCHResult), "结果应为GARCHResult类型"
+    
+    # 验证模型类型
+    assert "GARCH(1,1)" in result.model_type, "模型类型应为GARCH(1,1)"
+    
+    # 验证统计量合理性
+    assert result.n_obs == n, f"观测数量应为{n}"
+    assert len(result.coefficients) >= 3, "应至少有3个系数(omega, alpha, beta)"
+    
+    print("  模型类型:", result.model_type)
+    print("  系数数量:", len(result.coefficients))
+    print("  系数:", result.coefficients)
+    print("  对数似然值:", result.log_likelihood)
+    print("  AIC:", result.aic)
+    print("  持续性参数:", result.persistence)
+    print("  基本GARCH功能测试通过")
+
+
+def test_garch_order2():
+    """测试GARCH(2,1)模型"""
+    print("测试GARCH(2,1)模型...")
+    
+    # 生成测试数据（使用GARCH(1,1)数据）
+    np.random.seed(42)
+    n = 200
+    
+    # 生成简单的时间序列数据
+    data = []
+    for i in range(n):
+        data.append(np.random.randn() * 0.5)
+    
+    # 执行GARCH(2,1)模型
+    try:
+        result = garch_model(data, order=(2, 1))
+        
+        # 验证模型类型
+        assert "GARCH(2,1)" in result.model_type, "模型类型应为GARCH(2,1)"
+        
+        print("  模型类型:", result.model_type)
+        print("  GARCH(2,1)模型测试通过")
+    except Exception as e:
+        print(f"  GARCH(2,1)模型测试跳过（可能因数据量不足）: {e}")
+
+
+def test_garch_errors():
+    """测试GARCH错误处理"""
+    print("测试GARCH错误处理...")
+    
+    # 测试空数据
+    try:
+        garch_model([])
+        assert False, "应该抛出异常"
+    except Exception:
+        print("  空数据错误处理正确")
+    
+    # 测试无效阶数
+    try:
+        garch_model([1, 2, 3], order=(0, 0))
+        assert False, "应该抛出异常"
+    except Exception:
+        print("  无效阶数错误处理正确")
+    
+    print("  GARCH错误处理测试通过")
+
+
+if __name__ == "__main__":
+    print("开始测试GARCH模型...")
+    test_garch_basic()
+    test_garch_order2()
+    test_garch_errors()
+    print("所有GARCH测试通过!")

econometrics/tests/specific_data_modeling_tests/test_micro_discrete_limited_data.py

@@ -0,0 +1,189 @@
+"""
+微观离散与受限数据模型测试
+"""
+
+import numpy as np
+import pandas as pd
+import sys
+import os
+from scipy import stats
+
+# 添加项目根目录到路径
+sys.path.append(os.path.dirname(os.path.dirname(os.path.dirname(os.path.dirname(__file__)))))
+
+try:
+    from econometrics.specific_data_modeling.micro_discrete_limited_data import (
+        LogitModel,
+        ProbitModel,
+        TobitModel,
+        PoissonModel,
+        NegativeBinomialModel
+    )
+    HAS_MODELS = True
+except ImportError as e:
+    print(f"导入模型时出错: {e}")
+    HAS_MODELS = False
+
+
+def test_logit_model():
+    """测试Logit模型"""
+    print("测试Logit模型...")
+    
+    # 生成模拟数据
+    np.random.seed(42)
+    n = 1000
+    X = np.random.normal(0, 1, (n, 2))
+    coef_true = np.array([1.0, -0.5])
+    linear_pred = np.dot(X, coef_true)
+    prob = 1 / (1 + np.exp(-linear_pred))
+    y = np.random.binomial(1, prob)
+    
+    # 拟合模型
+    model = LogitModel()
+    model.fit(X, y)
+    
+    print(f"真实系数: {coef_true}")
+    print(f"估计系数: {model.results_.params[1:]}")  # 排除常数项
+    print(f"常数项: {model.results_.params[0]}")
+    print(f"AIC: {model.results_.aic}")
+    print(f"BIC: {model.results_.bic}")
+    print()
+
+
+def test_probit_model():
+    """测试Probit模型"""
+    print("测试Probit模型...")
+    
+    # 生成模拟数据
+    np.random.seed(42)
+    n = 1000
+    X = np.random.normal(0, 1, (n, 2))
+    coef_true = np.array([0.5, -0.3])
+    linear_pred = np.dot(X, coef_true)
+    prob = np.clip(stats.norm.cdf(linear_pred), 1e-10, 1-1e-10)
+    y = np.random.binomial(1, prob)
+    
+    # 拟合模型
+    model = ProbitModel()
+    model.fit(X, y)
+    
+    print(f"真实系数: {coef_true}")
+    print(f"估计系数: {model.results_.params[1:]}")  # 排除常数项
+    print(f"常数项: {model.results_.params[0]}")
+    print(f"AIC: {model.results_.aic}")
+    print(f"BIC: {model.results_.bic}")
+    print()
+
+
+def test_tobit_model():
+    """测试Tobit模型"""
+    print("测试Tobit模型...")
+    
+    # 生成模拟数据
+    np.random.seed(42)
+    n = 1000
+    X = np.random.normal(0, 1, (n, 2))
+    coef_true = np.array([1.0, -0.5])
+    sigma_true = 0.5
+    
+    # 生成潜在变量
+    y_latent = np.dot(X, coef_true) + np.random.normal(0, sigma_true, n)
+    
+    # 截断：低于0的值设为0
+    y = np.where(y_latent > 0, y_latent, 0)
+    
+    # 拟合模型
+    model = TobitModel(lower_bound=0)
+    model.fit(X, y)
+    
+    print(f"真实系数: {coef_true}")
+    print(f"估计系数: {model.results_.params[1:]}")  # 排除常数项
+    print(f"常数项: {model.results_.params[0]}")
+    print(f"对数似然: {model.results_.llf}")
+    print()
+
+
+def test_poisson_model():
+    """测试泊松模型"""
+    print("测试泊松模型...")
+    
+    # 生成模拟数据
+    np.random.seed(42)
+    n = 1000
+    X = np.random.normal(0, 1, (n, 2))
+    coef_true = np.array([0.5, -0.3])
+    mu = np.exp(np.dot(X, coef_true))
+    y = np.random.poisson(mu)
+    
+    # 拟合模型
+    model = PoissonModel()
+    model.fit(X, y)
+    
+    print(f"真实系数: {coef_true}")
+    print(f"估计系数: {model.results_.params[1:]}")  # 排除常数项
+    print(f"常数项: {model.results_.params[0]}")
+    print(f"对数似然: {model.results_.llf}")
+    print(f"AIC: {model.results_.aic}")
+    print(f"BIC: {model.results_.bic}")
+    
+    # 预测测试
+    y_pred = model.predict(X[:5])
+    print(f"前5个样本的预测值: {y_pred}")
+    print()
+
+
+def test_negative_binomial_model():
+    """测试负二项模型"""
+    print("测试负二项模型...")
+    
+    # 生成模拟数据
+    np.random.seed(42)
+    n = 1000
+    X = np.random.normal(0, 1, (n, 2))
+    coef_true = np.array([0.5, -0.3])
+    
+    mu = np.exp(np.dot(X, coef_true))
+    # 生成负二项分布数据
+    alpha = 0.5
+    size = 1.0 / alpha
+    prob = size / (size + mu)
+    y = np.random.negative_binomial(size, prob)
+    
+    # 拟合模型
+    model = NegativeBinomialModel()
+    model.fit(X, y)
+    
+    print(f"真实系数: {coef_true}")
+    print(f"估计系数: {model.results_.params[1:-1]}")  # 排除常数项和alpha参数
+    print(f"常数项: {model.results_.params[0]}")
+    print(f"对数似然: {model.results_.llf}")
+    print(f"AIC: {model.results_.aic}")
+    print(f"BIC: {model.results_.bic}")
+    
+    # 预测测试
+    y_pred = model.predict(X[:5])
+    print(f"前5个样本的预测值: {y_pred}")
+    print()
+
+
+if __name__ == "__main__":
+    if not HAS_MODELS:
+        print("模型不可用，请确保已安装statsmodels库")
+        exit(1)
+        
+    try:
+        import statsmodels.api as sm
+    except ImportError:
+        print("需要安装statsmodels库")
+        exit(1)
+    
+    print("微观离散与受限数据模型测试")
+    print("=" * 50)
+    
+    test_logit_model()
+    test_probit_model()
+    test_tobit_model()
+    test_poisson_model()
+    test_negative_binomial_model()
+    
+    print("所有测试完成！")

econometrics/tests/specific_data_modeling_tests/test_unit_root.py

@@ -0,0 +1,156 @@
+"""
+单位根检验测试脚本
+"""
+
+import sys
+import os
+import numpy as np
+
+# 添加项目根目录到路径
+sys.path.append(os.path.join(os.path.dirname(__file__), '..', '..', '..'))
+
+from econometrics.specific_data_modeling.time_series_panel_data.unit_root_tests import adf_test, pp_test, kpss_test, UnitRootTestResult
+
+
+def test_adf_test():
+    """测试ADF单位根检验"""
+    print("测试ADF单位根检验...")
+    
+    # 生成平稳时间序列数据 (AR(1)过程，系数<1)
+    np.random.seed(42)
+    n = 100
+    ar_param = 0.6  # |ar_param| < 1，序列平稳
+    data = [0]  # 初始值
+    
+    # 生成AR(1)过程: y[t] = 0.6 * y[t-1] + noise
+    for i in range(1, n):
+        data.append(ar_param * data[i-1] + np.random.randn() * 0.5)
+    
+    # 执行ADF检验
+    result = adf_test(data)
+    
+    # 验证结果类型
+    assert isinstance(result, UnitRootTestResult), "结果应为UnitRootTestResult类型"
+    
+    # 验证检验类型
+    assert "Augmented Dickey-Fuller" in result.test_type, "检验类型应为ADF"
+    
+    # 验证统计量合理性
+    assert result.n_obs == n, f"观测数量应为{n}"
+    assert isinstance(result.stationary, bool), "平稳性判断应为布尔值"
+    
+    print("  检验统计量:", result.test_statistic)
+    print("  p值:", result.p_value)
+    print("  是否平稳:", result.stationary)
+    print("  ADF单位根检验测试通过")
+
+
+def test_adf_test_nonstationary():
+    """测试ADF单位根检验（非平稳序列）"""
+    print("测试ADF单位根检验（非平稳序列）...")
+    
+    # 生成非平稳时间序列数据 (随机游走)
+    np.random.seed(42)
+    n = 100
+    data = [0]  # 初始值
+    
+    # 生成随机游走过程: y[t] = y[t-1] + noise
+    for i in range(1, n):
+        data.append(data[i-1] + np.random.randn() * 0.3)
+    
+    # 执行ADF检验
+    result = adf_test(data)
+    
+    print("  检验统计量:", result.test_statistic)
+    print("  p值:", result.p_value)
+    print("  是否平稳:", result.stationary)
+    print("  ADF单位根检验（非平稳序列）测试通过")
+
+
+def test_pp_test():
+    """测试PP单位根检验"""
+    print("测试PP单位根检验...")
+    
+    # 生成平稳时间序列数据
+    np.random.seed(42)
+    n = 100
+    ar_param = 0.7
+    data = [0]  # 初始值
+    
+    # 生成AR(1)过程: y[t] = 0.7 * y[t-1] + noise
+    for i in range(1, n):
+        data.append(ar_param * data[i-1] + np.random.randn() * 0.5)
+    
+    # 执行PP检验
+    result = pp_test(data)
+    
+    # 验证结果类型
+    assert isinstance(result, UnitRootTestResult), "结果应为UnitRootTestResult类型"
+    
+    # 验证检验类型
+    assert "Phillips-Perron" in result.test_type, "检验类型应为PP"
+    
+    print("  检验统计量:", result.test_statistic)
+    print("  p值:", result.p_value)
+    print("  是否平稳:", result.stationary)
+    print("  PP单位根检验测试通过")
+
+
+def test_kpss_test():
+    """测试KPSS单位根检验"""
+    print("测试KPSS单位根检验...")
+    
+    # 生成平稳时间序列数据
+    np.random.seed(42)
+    n = 100
+    data = np.random.randn(n).tolist()  # 白噪声序列，平稳
+    
+    # 执行KPSS检验
+    result = kpss_test(data)
+    
+    # 验证结果类型
+    assert isinstance(result, UnitRootTestResult), "结果应为UnitRootTestResult类型"
+    
+    # 验证检验类型
+    assert "KPSS" in result.test_type, "检验类型应为KPSS"
+    
+    print("  检验统计量:", result.test_statistic)
+    print("  p值:", result.p_value)
+    print("  是否平稳:", result.stationary)
+    print("  KPSS单位根检验测试通过")
+
+
+def test_unit_root_errors():
+    """测试单位根检验错误处理"""
+    print("测试单位根检验错误处理...")
+    
+    # 测试空数据
+    try:
+        adf_test([])
+        assert False, "应该抛出异常"
+    except Exception:
+        print("  ADF空数据错误处理正确")
+    
+    try:
+        pp_test([])
+        assert False, "应该抛出异常"
+    except Exception:
+        print("  PP空数据错误处理正确")
+    
+    try:
+        kpss_test([])
+        assert False, "应该抛出异常"
+    except Exception:
+        print("  KPSS空数据错误处理正确")
+    
+    print("  单位根检验错误处理测试通过")
+
+
+if __name__ == "__main__":
+    print("开始测试单位根检验...")
+    test_adf_test()
+    test_adf_test_nonstationary()
+    test_pp_test()
+    test_kpss_test()
+    test_unit_root_errors()
+    print("所有单位根检验测试通过!")

econometrics/tests/specific_data_modeling_tests/test_var.py

@@ -0,0 +1,124 @@
+"""
+VAR模型测试脚本
+"""
+
+import sys
+import os
+import numpy as np
+
+# 添加项目根目录到路径
+sys.path.append(os.path.join(os.path.dirname(__file__), '..', '..', '..'))
+
+from econometrics.specific_data_modeling.time_series_panel_data.var_svar_model import var_model, svar_model, VARResult
+
+
+def test_var_basic():
+    """测试基本VAR功能"""
+    print("测试基本VAR功能...")
+    
+    # 生成VAR(1)过程数据
+    np.random.seed(42)
+    n = 100
+    
+    # VAR(1)系数矩阵
+    A = np.array([[0.5, 0.2], 
+                  [0.1, 0.3]])
+    
+    # 生成VAR过程
+    data = [np.array([0.0, 0.0])]  # 初始值
+    
+    for i in range(1, n):
+        # VAR(1): y[t] = A @ y[t-1] + noise
+        noise = np.random.multivariate_normal([0, 0], [[0.5, 0.1], [0.1, 0.5]])
+        new_value = A @ data[i-1] + noise
+        data.append(new_value)
+    
+    # 转换为列表格式
+    data = np.array(data).T  # 转置以匹配要求的格式
+    data_list = [data[0].tolist(), data[1].tolist()]
+    
+    # 执行VAR模型
+    result = var_model(data_list, lags=1, variables=['y1', 'y2'])
+    
+    # 验证结果类型
+    assert isinstance(result, VARResult), "结果应为VARResult类型"
+    
+    # 验证模型类型
+    assert "VAR(1)" in result.model_type, "模型类型应为VAR(1)"
+    
+    # 验证统计量合理性
+    assert result.n_obs == n, f"观测数量应为{n}"
+    assert len(result.variables) == 2, "变量数量应为2"
+    
+    print("  模型类型:", result.model_type)
+    print("  变量:", result.variables)
+    print("  AIC:", result.aic)
+    print("  基本VAR功能测试通过")
+
+
+def test_var_lag2():
+    """测试2阶滞后的VAR模型"""
+    print("测试2阶滞后的VAR模型...")
+    
+    # 生成VAR(2)过程数据
+    np.random.seed(42)
+    n = 120
+    
+    # 生成VAR过程
+    data = [np.array([0.0, 0.0])]  # 初始值
+    data.append(np.array([0.1, 0.05]))  # 第二个初始值
+    
+    # VAR(2)系数矩阵
+    A1 = np.array([[0.5, 0.2], 
+                   [0.1, 0.3]])
+    A2 = np.array([[0.1, -0.1], 
+                   [0.05, 0.2]])
+    
+    for i in range(2, n):
+        # VAR(2): y[t] = A1 @ y[t-1] + A2 @ y[t-2] + noise
+        noise = np.random.multivariate_normal([0, 0], [[0.5, 0.1], [0.1, 0.5]])
+        new_value = A1 @ data[i-1] + A2 @ data[i-2] + noise
+        data.append(new_value)
+    
+    # 转换为列表格式
+    data = np.array(data).T  # 转置以匹配要求的格式
+    data_list = [data[0].tolist(), data[1].tolist()]
+    
+    # 执行VAR模型 (滞后2阶)
+    result = var_model(data_list, lags=2, variables=['y1', 'y2'])
+    
+    # 验证模型类型
+    assert "VAR(2)" in result.model_type, "模型类型应为VAR(2)"
+    
+    print("  模型类型:", result.model_type)
+    print("  2阶滞后的VAR模型测试通过")
+
+
+def test_var_errors():
+    """测试VAR错误处理"""
+    print("测试VAR错误处理...")
+    
+    # 测试空数据
+    try:
+        var_model([])
+        assert False, "应该抛出异常"
+    except Exception:
+        print("  空数据错误处理正确")
+    
+    # 测试不一致的序列长度
+    try:
+        data_list = [[1, 2, 3], [1, 2]]  # 不一致的长度
+        var_model(data_list)
+        assert False, "应该抛出异常"
+    except Exception:
+        print("  不一致序列长度错误处理正确")
+    
+    print("  VAR错误处理测试通过")
+
+
+if __name__ == "__main__":
+    print("开始测试VAR模型...")
+    test_var_basic()
+    test_var_lag2()
+    test_var_errors()
+    print("所有VAR测试通过!")