aigroup-econ-mcp 0.4.2__py3-none-any.whl → 1.4.3__py3-none-any.whl

aigroup_econ_mcp-1.4.3.dist-info/RECORD

@@ -0,0 +1,92 @@
+./.gitignore,sha256=M67x66mQGPjpk5A6LO93G9R7cY9vWCQH9sdLFt6S53c,3284
+./LICENSE,sha256=DoyCJUWlDzKbqc5KRbFpsGYLwLh-XJRHKQDoITjb1yc,1083
+./PKG-INFO,sha256=RcxXpD22GnkhG0mk9c2aCVZ41xwTp-ZzI4Ow6-u4ER4,23096
+./README.md,sha256=Yyb3U92_cxxEmZuSZNLDHThEwlZwBV81xMu3FqVCkbA,22067
+./__init__.py,sha256=4R069rUFbLRbjGo_pn8C-R330y3dXz6HxdELunjTTok,282
+./cli.py,sha256=1pppkNYcSj0-bmYSpwx6TBVI0qC1bw8H5_vRlwgQV2M,639
+./pyproject.toml,sha256=vkAKNgc72ZzYaujDMsLP14FoX20SxTZPgeCQV_RN4q8,2139
+./server.py,sha256=OBTiYiG6XExQhgtJXmahnCeoL0yBGTS0QX1F9Se8Lcw,2542
+./econometrics/README.md,sha256=kbBTUhv_ZHluD6ultI0FqSGqBKNc95OmaghESyrT0ZM,5499
+./econometrics/__init__.py,sha256=PXN3is7i4PU4YVX4ZcDqrAOWEiizjA13TPQb68i47OM,4328
+./econometrics/advanced_methods/modern_computing_machine_learning/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/basic_parametric_estimation/__init__.py,sha256=IFgOw46qAzecWUha-uUlab0VptCh82Mhc1hsZIlNH8I,463
+./econometrics/basic_parametric_estimation/gmm/__init__.py,sha256=H5KwHqEx-ALHNlI4DyOenj4XEJ5PI-vQ0kAXR9o_S1s,151
+./econometrics/basic_parametric_estimation/gmm/gmm_model.py,sha256=uwBIxjmO58P41qx7lHSkECMh3WhQoZgYH4gyIDLnwQI,9410
+./econometrics/basic_parametric_estimation/mle/__init__.py,sha256=w9eJFI5c-pcMqG7igIFedz5FOnQGIviZiXXc4Dso_-s,154
+./econometrics/basic_parametric_estimation/mle/mle_model.py,sha256=sQcPlheEYD5Kpn0RDNKjy305AMwCP07kK2U6FN73u_Q,8526
+./econometrics/basic_parametric_estimation/ols/__init__.py,sha256=NhFFi8COkCRrSJrcjvEVOlSJkQABa20PO9yFpykJ4X4,157
+./econometrics/basic_parametric_estimation/ols/ols_model.py,sha256=vj98YSotw-7EbQMzGWG3CGRsgpI2C4axr-KgmmvmxcE,4655
+./econometrics/causal_inference/causal_identification_strategy/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/missing_data/missing_data_measurement_error/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/model_specification_diagnostics_robust_inference/README.md,sha256=A6MQ3e_fBRcgUTGG-8PIZoak2mKxg8wvoOo_mF52guk,4681
+./econometrics/model_specification_diagnostics_robust_inference/__init__.py,sha256=EmJt_7fLMK3d6MkNCWYhWLPsZgHjjIj0zs9xkM9nh38,1832
+./econometrics/model_specification_diagnostics_robust_inference/diagnostic_tests/__init__.py,sha256=KRhdc5K-xATvNQPnSBRnLRr3Hhi2mtbm6GG2_dYdsBU,446
+./econometrics/model_specification_diagnostics_robust_inference/diagnostic_tests/diagnostic_tests_model.py,sha256=xqagx4CcqLbyIwJXcmnnb33TWe_0tXSwtqCUCIVHtZo,5119
+./econometrics/model_specification_diagnostics_robust_inference/generalized_least_squares/__init__.py,sha256=3ELQkgGtOLgWY6UOi_ZFL3g7KQxid1WCMdhhByoZHUg,196
+./econometrics/model_specification_diagnostics_robust_inference/generalized_least_squares/gls_model.py,sha256=HupVc4WkpltRCkEZYYcrdJG_8cdY7p6SJS-_v6emKqk,4418
+./econometrics/model_specification_diagnostics_robust_inference/model_selection/__init__.py,sha256=kDxpczAVXNaam7T_Ri0SDKl42eY4hzzu8M_XzVt-M-4,319
+./econometrics/model_specification_diagnostics_robust_inference/model_selection/model_selection_model.py,sha256=V0gTaHjIl_xRWAfILLIn-Z0Qt12E8Q2UMQwBVvsihgY,9297
+./econometrics/model_specification_diagnostics_robust_inference/regularization/__init__.py,sha256=48Y9Lhm77kMtGtH-K1hrpEf41cNf10dSQO5As3Mb2RU,311
+./econometrics/model_specification_diagnostics_robust_inference/regularization/regularization_model.py,sha256=kzkhAEN6pGSk3US-o4_vh2CwhcBpBoNTbAaDBNCD9P8,6047
+./econometrics/model_specification_diagnostics_robust_inference/robust_errors/__init__.py,sha256=J-PaSUus5_f3t9kSC9EGWZPYcxwwCU5KEzymDYQi7ZA,257
+./econometrics/model_specification_diagnostics_robust_inference/robust_errors/robust_errors_model.py,sha256=ctOQGLSRXCdKbReUw6WyjlFhHsnrl7RfLkmI92U-2tk,4269
+./econometrics/model_specification_diagnostics_robust_inference/simultaneous_equations/__init__.py,sha256=gQg1Hjt3QWuLu8nPqUDG3GbsEmL9U31LuRpLAb2CxLI,294
+./econometrics/model_specification_diagnostics_robust_inference/simultaneous_equations/simultaneous_equations_model.py,sha256=LQbOf6o6iVtcrJBgpi6txznaxtn9REBQCIgzBAZnlxM,10309
+./econometrics/model_specification_diagnostics_robust_inference/weighted_least_squares/__init__.py,sha256=uUcFyCn7Ny3jkAw7GsB2UYT_ucrsbM1P6wGx7MAWoRc,193
+./econometrics/model_specification_diagnostics_robust_inference/weighted_least_squares/wls_model.py,sha256=Jzc5p3TI1elFuYQp0EgGvTpsgtBiqKmKAPo5ppZ7xq0,4196
+./econometrics/nonparametric/nonparametric_semiparametric_methods/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/spatial_econometrics/spatial_econometrics_new/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/specific_data_modeling/micro_discrete_limited_data/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/specific_data_modeling/survival_duration_data/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/specific_data_modeling/time_series_panel_data/__init__.py,sha256=HGRg4teWO8cbzr4oS_Zx2iqNJppAxRPp7WsaAt6CqMY,2526
+./econometrics/specific_data_modeling/time_series_panel_data/arima_model.py,sha256=F_uhdZ0UsKCfGrJ1WscLmiyF4OAFKtqz_OJw5My7OYU,4162
+./econometrics/specific_data_modeling/time_series_panel_data/cointegration_vecm.py,sha256=hGxOlGOchQQIEkzds8sHWlUrnenObJKjC1DctJx9jLQ,14048
+./econometrics/specific_data_modeling/time_series_panel_data/dynamic_panel_models.py,sha256=a_d-F5Vtv5XCpshOiduvICeBTzGbtx5f4wOqFLzM1Xo,28094
+./econometrics/specific_data_modeling/time_series_panel_data/exponential_smoothing.py,sha256=_EMoy13MfgZl2xNIZOHVOisihAvSsQo9RCs6ob38mnE,7354
+./econometrics/specific_data_modeling/time_series_panel_data/garch_model.py,sha256=1-cnQz8qV6kPHJgo61g0rmF8cTKArqaMJPX-1o24Y3Y,7033
+./econometrics/specific_data_modeling/time_series_panel_data/panel_diagnostics.py,sha256=zo2mkWvL8e4lIuQ_UT25geMnZq_uUPuiIIyuoIyCtnI,3910
+./econometrics/specific_data_modeling/time_series_panel_data/panel_var.py,sha256=fzHkwORKwrCFEY24aExTGTdYXdtaiv_fZUJ-anySS5g,2225
+./econometrics/specific_data_modeling/time_series_panel_data/structural_break_tests.py,sha256=lBJMSzvMG8KFuKBqjOYM_lPD_P_LXEBqnBmYRiaNJjk,2362
+./econometrics/specific_data_modeling/time_series_panel_data/time_varying_parameter_models.py,sha256=iXGMdhcfPouDYa9VUQrmmAcjwvZA4AatKPxpFapvEes,3546
+./econometrics/specific_data_modeling/time_series_panel_data/unit_root_tests.py,sha256=K6aTAYTF73Mcg7pLBK3tDY2wWsfXSc1xphOWBm2EfQM,6854
+./econometrics/specific_data_modeling/time_series_panel_data/var_svar_model.py,sha256=1IviPC5XkQBsEVI09poonRXq5RAVDmK6Lu5pvsrgPIM,14742
+./econometrics/statistical_inference/statistical_inference_techniques/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/statistics/distribution_decomposition_methods/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./econometrics/tests/basic_parametric_estimation_tests/__init__.py,sha256=m4Nv29MZh4w-TekKYMdP5hLPxfIvw_X3s1a9fdnxjQk,53
+./econometrics/tests/basic_parametric_estimation_tests/test_gmm.py,sha256=BA6KMQaYRLxnWJgw-LclCnJu_-npAiDtom5LsAAozNM,3760
+./econometrics/tests/basic_parametric_estimation_tests/test_mle.py,sha256=kZa3sx7Cdu7s6hrx645AKXJNN7hnRJantw0VgBo_Mis,3735
+./econometrics/tests/basic_parametric_estimation_tests/test_ols.py,sha256=dqsrYhmjKcwKVdRdhBHPaygukvsZ_mFsj4y7BoxJgCI,2835
+./econometrics/tests/model_specification_diagnostics_tests/__init__.py,sha256=r_fvo5IW5smEMykzU0LGdd8vtL3NOI2ZyL4WgVOoYak,71
+./econometrics/tests/model_specification_diagnostics_tests/test_diagnostic_tests.py,sha256=7ICFo3rDo2yK9rGcWtYSNe1iIEZO6ywp4oYFHVL_z4o,2694
+./econometrics/tests/model_specification_diagnostics_tests/test_robust_errors.py,sha256=c5C7VcsvNmegIAzsok4q5iJKb5Lstpa8qHdY7jdW0g4,2905
+./econometrics/tests/specific_data_modeling_tests/__init__.py,sha256=pT2dhJ9FeaPzTY4xcwuMYCk6rtnYie_xTwk7LARuFg4,53
+./econometrics/tests/specific_data_modeling_tests/test_arima.py,sha256=aQGhShPzDdoDjU5GociZDJk190OhLbD1A2JRXIjDv6A,2806
+./econometrics/tests/specific_data_modeling_tests/test_dynamic_panel.py,sha256=99FEc_o437W59Mfj8jzlb2m7QDCCWANWwMN0aEL24IU,6372
+./econometrics/tests/specific_data_modeling_tests/test_exponential_smoothing.py,sha256=D90bZhBfDAIcT1IBhjkRluv0xpxJsulzAEudI4FLjKI,3366
+./econometrics/tests/specific_data_modeling_tests/test_garch.py,sha256=Fkaih4Lh5TVScCBKjtXI60pSxnobi2aaGcca8p04PM4,3271
+./econometrics/tests/specific_data_modeling_tests/test_unit_root.py,sha256=87gEu8fDh7LS20xMJk_OrBryA2UYa-8KPUBaVaNFjNc,4539
+./econometrics/tests/specific_data_modeling_tests/test_var.py,sha256=kbWZVftblOULYABBBCiIS5x6JpBdPv-ZgG5HfbtsI4U,3556
+./prompts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./prompts/analysis_guides.py,sha256=FCsEtRCETWJUuG8exmAL30NRpD0eT9dyqljlpxXU5u0,1545
+./resources/MCP_MASTER_GUIDE.md,sha256=lZ2R2xV1ksCHSkSl4YorvW0ZmgCJM3Oq8UuM0kZaG6I,13407
+./resources/MCP_TOOLS_DATA_FORMAT_GUIDE.md,sha256=LFY_VS_bwLu5LBi1vbPruFac4qwGzP-UwvDAk1s9ha0,5706
+./resources/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+./tools/README.md,sha256=KkEUtBa3TBwz2HcVoW1Z9sGExb8BrElKzcgSlG0eoP0,2217
+./tools/__init__.py,sha256=y9A19bfUGDzDcS2BySRrGt8QhdH9_Ahk5s5g4eiURRI,1075
+./tools/data_loader.py,sha256=-HFKORi03DejxCBD2i5HJJCiwEEdG5uSA4jJivhVT0Y,6886
+./tools/decorators.py,sha256=ascsn0rNtzv1-RZWa4xQXArFD-43R3YuUbVIyxRtuOA,867
+./tools/econometrics_adapter.py,sha256=MHXvzjBnULEgBGc5HRV5m93auzwPhSZyorl8dq3Mt_0,11429
+./tools/mcp_tools_registry.py,sha256=0st55db2HWxSx2VWeyY89h_PkP_ULQQ7ydBg0ZyRNr0,3462
+./tools/model_specification_adapter.py,sha256=M1bdELqKJe7Mg91V7BNqdBXV2D5l3nqC_h91sQ4Ifno,14570
+./tools/output_formatter.py,sha256=l03L6QhsSKiOoDBSDeMkRjkxVHvfMAuBFcTdB7-m2CM,20751
+./tools/time_series_panel_data_adapter.py,sha256=b9GsTzLIEQnUInC4jVUasOQbEmiEvhy60QxAWgznSWs,35746
+./tools/time_series_panel_data_tools.py,sha256=W4R_CRBYnRjAd5s3_6_BfOPuIluxTkfj7iAfH-UWBgs,2098
+./tools/mcp_tool_groups/__init__.py,sha256=5kVm7UKVJTdj8YWA6bhODUaXgONdEjAGjy31gVnry5o,21
+./tools/mcp_tool_groups/basic_parametric_tools.py,sha256=9noEMDxQi7TfDlSXEl5RYEEd1RmQYT-FfzZtmjUUbIw,5827
+./tools/mcp_tool_groups/model_specification_tools.py,sha256=-IqFF7OjXKwKaFU-XruOFO62AkvEVJqSHHgh8ymSj8s,14834
+./tools/mcp_tool_groups/time_series_tools.py,sha256=IytweixTmJ14uMuAFczZ80mV6eUBrF85AQrBUkmpcCA,18276
+aigroup_econ_mcp-1.4.3.dist-info/METADATA,sha256=RcxXpD22GnkhG0mk9c2aCVZ41xwTp-ZzI4Ow6-u4ER4,23096
+aigroup_econ_mcp-1.4.3.dist-info/WHEEL,sha256=qtCwoSJWgHk21S1Kb4ihdzI2rlJ1ZKaIurTj_ngOhyQ,87
+aigroup_econ_mcp-1.4.3.dist-info/entry_points.txt,sha256=UbLsgggIXKYqolHc0xxLCKr0_aFxRek4_QOCAgzDHZk,45
+aigroup_econ_mcp-1.4.3.dist-info/licenses/LICENSE,sha256=DoyCJUWlDzKbqc5KRbFpsGYLwLh-XJRHKQDoITjb1yc,1083
+aigroup_econ_mcp-1.4.3.dist-info/RECORD,,

aigroup_econ_mcp-1.4.3.dist-info/entry_points.txt

@@ -0,0 +1,2 @@
+[console_scripts]
+aigroup-econ-mcp = cli:cli

aigroup_econ_mcp-1.4.3.dist-info/licenses/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 AIGroup
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

cli.py

@@ -0,0 +1,28 @@
+#!/usr/bin/env python3
+"""
+AIGroup Econometrics MCP - CLI 入口
+"""
+
+import sys
+import os
+from pathlib import Path
+
+def main():
+    """CLI 主函数"""
+    try:
+        # 导入并运行服务器
+        from server import main as server_main
+        server_main()
+    except ImportError as e:
+        print(f"导入错误: {e}")
+        print("请确保所有依赖已正确安装")
+        sys.exit(1)
+    except Exception as e:
+        print(f"启动服务器时出错: {e}")
+        sys.exit(1)
+
+# 添加cli函数以匹配pyproject.toml中的入口点定义
+cli = main
+
+if __name__ == "__main__":
+    main()

econometrics/README.md

@@ -0,0 +1,18 @@
+# 计量经济学模型说明文档
+
+本目录包含AIGroup计量经济学MCP服务器的所有计量经济学分析功能。
+| 核心类别 | 解决的核心问题 | 主要方法 |
+| :--- | :--- | :--- |
+| **1. 基础与参数估计** | 建立变量间的基础参数化关系，并进行估计。 | **普通最小二乘法 (OLS)**、**最大似然估计 (MLE)**、**广义矩估计 (GMM)** |
+| **2. 模型设定、诊断与稳健推断** | 当基础模型的理想假设不成立时，修正模型或调整推断；对模型进行诊断和选择。 | **稳健标准误**（处理异方差/自相关）、**广义最小二乘法 (GLS)**、**加权最小二乘法 (WLS)**、**岭回归/LASSO/弹性网络**（处理多重共线性/高维数据）、**联立方程模型**（处理双向因果关系）<br><br>**模型诊断**：异方差检验（White、Breusch-Pagan）、自相关检验（Durbin-Watson、Ljung-Box）、正态性检验（Jarque-Bera）、多重共线性诊断（VIF）、内生性检验（Durbin-Wu-Hausman）、残差诊断、影响点分析<br><br>**模型选择**：信息准则（AIC/BIC/HQIC）、交叉验证（K折、留一法）、格兰杰因果检验 |
+| **3. 因果识别策略** | 在非实验数据中，识别变量间的因果关系（解决内生性问题）。 | **工具变量法 (IV/2SLS)**、**控制函数法**、**面板数据模型**（固定效应、随机效应、一阶差分、Hausman检验）、**双重差分法 (DID)**、**三重差分法 (DDD)**、**事件研究法 (Event Study)**、**断点回归设计 (RDD)**、**合成控制法**、**匹配方法**（倾向得分匹配PSM、协变量平衡、倾向得分加权IPW、熵平衡法）<br><br>**效应分解与异质性**：中介效应分析（Baron-Kenny、Bootstrap检验、Sobel检验）、调节效应分析（交互项回归）、处理效应异质性 (HTE)、条件平均处理效应 (CATE)、因果森林<br><br>**稳健性检验**：敏感性分析、Rosenbaum bounds、双重机器学习 (Double ML) |
+| **4. 特定数据类型建模** | 针对因变量或数据结构的固有特性进行建模。 | |
+| &nbsp;&nbsp;**• 微观离散与受限数据** | 因变量为分类、计数、截断等非连续情况。 | **Logit/Probit**、**多项/有序/条件Logit**、**混合/嵌套Logit**、**Tobit**、**泊松/负二项回归**、**Heckman选择模型** |
+| &nbsp;&nbsp;**• 时间序列与面板数据** | 分析具有时间维度数据的动态依赖、预测和非平稳性。 | **ARIMA**、**指数平滑法**、**VAR/SVAR**、**GARCH**、**协整分析/VECM**、**面板VAR**<br><br>**平稳性与单位根检验**：ADF检验、PP检验、KPSS检验<br><br>**动态面板模型**：Arellano-Bond估计（差分GMM）、Blundell-Bond估计（系统GMM）<br><br>**结构突变检验**：Chow检验、Quandt-Andrews检验、Bai-Perron检验（多重断点）<br><br>**面板数据诊断**：Hausman检验（FE vs RE）、F检验（Pooled vs FE）、LM检验（Pooled vs RE）、组内相关性检验<br><br>**时变参数模型**：门限模型/转换回归（TAR/STAR）、马尔科夫转换模型 |
+| &nbsp;&nbsp;**• 生存/持续时间数据** | 分析"事件发生时间"数据并处理右删失。 | **Kaplan-Meier估计量**、**Cox比例风险模型**、**加速失效时间模型** |
+| **5. 空间计量经济学** | 处理数据的空间依赖性和空间异质性。 | **空间权重矩阵构建**（邻接、距离、K近邻矩阵）<br><br>**空间自相关检验**：Moran's I、Geary's C、局部空间自相关 (LISA)<br><br>**空间回归模型**：空间滞后模型 (SAR)、空间误差模型 (SEM)、空间杜宾模型 (SDM)、地理加权回归 (GWR)、空间面板数据模型 |
+| **6. 非参数与半参数方法** | 放宽函数形式的线性或参数化假设，让数据本身驱动关系形态。 | **核回归**、**局部回归**、**样条回归**、**广义可加模型 (GAM)**、**部分线性模型**、**非参数工具变量估计** |
+| **7. 分布分析与分解方法** | 分析因变量整个条件分布的特征，而非仅仅条件均值；对差异或变化进行分解。 | **分位数回归**<br><br>**分解方法**：Oaxaca-Blinder分解、DiNardo-Fortin-Lemieux反事实分解、方差分解、ANOVA分解、Shapley值分解、时间序列分解（趋势-季节-随机） |
+| **8. 现代计算与机器学习** | 处理高维数据、复杂模式识别、预测以及为因果推断提供辅助工具。 | **监督学习**：随机森林、梯度提升机 (GBM/XGBoost)、支持向量机 (SVM)、神经网络<br><br>**无监督学习**：聚类分析（K-means、层次聚类）<br><br>**因果推断增强**：双重机器学习 (Double ML)、因果森林 (Causal Forest) |
+| **9. 统计推断技术** | 在理论分布难以推导或模型复杂时，进行可靠的区间估计与假设检验。 | **重采样方法**：自助法 (Bootstrap)、Pairs Bootstrap、Residual Bootstrap、Wild Bootstrap（异方差）、Block Bootstrap（时间序列/面板）、刀切法 (Jackknife)<br><br>**模拟方法**：蒙特卡洛模拟、置换检验 (Permutation Test)<br><br>**渐近方法**：Delta方法、聚类稳健推断 |
+| **10. 缺失数据与测量误差** | 处理数据不完整或变量测量不准确的问题。 | **缺失数据处理**：列表删除法、均值插补、回归插补、多重插补 (Multiple Imputation - MICE/Amelia)、期望最大化算法 (EM)<br><br>**测量误差**：工具变量法、SIMEX方法 |

econometrics/__init__.py

@@ -0,0 +1,191 @@
+# 基础与参数估计模块
+from .basic_parametric_estimation import (
+    OLSResult,
+    ols_regression,
+    MLEResult,
+    mle_estimation,
+    GMMResult,
+    gmm_estimation
+)
+
+# 二元选择模型模块
+# from .discrete_choice.binary_choice import (
+#     logit_model,
+#     probit_model,
+#     BinaryChoiceResult
+# )
+
+# 多项选择模型模块
+# from .discrete_choice.multinomial_choice import (
+#     multinomial_logit,
+#     ordered_choice_model,
+#     MultinomialResult,
+#     OrderedResult
+# )
+
+# 计数数据模型模块
+# from .discrete_choice.count_data_models import (
+#     poisson_regression,
+#     negative_binomial_regression,
+#     tobit_model,
+#     PoissonResult,
+#     NegativeBinomialResult,
+#     TobitResult
+# )
+
+# 非参数回归模块
+# from .nonparametric.nonparametric_regression import (
+#     kernel_regression,
+#     local_polynomial_regression,
+#     NonparametricRegressionResult
+# )
+
+# 样条和GAM模块
+# from .nonparametric.spline_gam import (
+#     spline_regression,
+#     generalized_additive_model,
+#     SplineResult,
+#     GAMResult
+# )
+
+# 条件期望函数
+# from .nonparametric.conditional_expectation_functions import (
+#     conditional_expectation_function,
+#     CEFResult
+# )
+
+# 面板数据分析模块
+# from .panel_data.panel_data_models import (
+#     fixed_effects_model,
+#     random_effects_model,
+#     PanelDataResult
+# )
+
+# from .panel_data.panel_unit_root_tests import (
+#     levin_lin_test,
+#     im_pesaran_shin_test,
+#     madwu_test
+# )
+
+# 时间序列分析模块
+# from .time_series.time_series_models import (
+#     ar_model,
+#     arma_model,
+#     var_model,
+#     TimeSeriesResult
+# )
+
+# from .time_series.advanced_time_series import (
+#     garch_model,
+#     state_space_model,
+#     variance_decomposition
+# )
+
+# 高级计量方法模块
+# from .advanced_methods.advanced_methods import (
+#     psm_model,
+#     did_model,
+#     rdd_model,
+#     AdvancedMethodsResult
+# )
+
+# from .advanced_methods.quantile_regression import (
+#     quantile_regression,
+#     QuantileRegressionResult
+# )
+
+# from .advanced_methods.survival_analysis import (
+#     cox_model,
+#     kaplan_meier_estimation,
+#     SurvivalAnalysisResult
+# )
+
+# 统计推断模块
+# from .statistical_inference.hypothesis_testing import (
+#     t_test,
+#     f_test,
+#     chi2_test,
+#     HypothesisTestResult
+# )
+
+# from .statistical_inference.confidence_intervals import (
+#     confidence_interval,
+#     ConfidenceIntervalResult
+# )
+
+# from .statistical_inference.bootstrapping import (
+#     bootstrap_inference,
+#     BootstrapResult
+# )
+
+# 模型设定、诊断和稳健推断模块
+# from .model_specification_diagnostics_robust_inference.model_specification import (
+#     reset_test,
+#     ModelSpecificationResult
+# )
+
+# from .model_specification_diagnostics_robust_inference.model_diagnostics import (
+#     heteroskedasticity_test,
+#     autocorrelation_test,
+#     ModelDiagnosticsResult
+# )
+
+# from .model_specification_diagnostics_robust_inference.robust_inference import (
+#     robust_se_estimation,
+#     RobustInferenceResult
+# )
+
+# 缺失数据处理模块
+# from .missing_data.missing_data_methods import (
+#     multiple_imputation,
+#     inverse_probability_weighting,
+#     MissingDataResult
+# )
+
+# 因果推断模块
+# from .causal_inference.causal_inference_methods import (
+#     instrumental_variables,
+#     regression_discontinuity,
+#     CausalInferenceResult
+# )
+
+# 空间计量经济学模块
+# from .spatial_econometrics.spatial_econometrics import (
+#     spatial_lag_model,
+#     spatial_error_model,
+#     SpatialEconometricsResult
+# )
+
+# 特定数据建模模块
+# from .specific_data_modeling.heterogeneous_data_models import (
+#     mixed_effects_model,
+#     HeterogeneousDataResult
+# )
+
+# from .specific_data_modeling.micro_discrete_limited_data import (
+#     tobit_model_2,
+#     MicroDiscreteDataResult
+# )
+
+# from .specific_data_modeling.censored_truncated_data import (
+#     tobit_model_3,
+#     truncated_regression,
+#     CensoredTruncatedDataResult
+# )
+
+# 异常类
+# from .exceptions import (
+#     EconometricToolError,
+#     DataValidationError,
+#     ModelFittingError,
+#     ConfigurationError
+# )
+
+__all__ = [
+    "OLSResult",
+    "ols_regression",
+    "MLEResult", 
+    "mle_estimation",
+    "GMMResult",
+    "gmm_estimation"
+]

econometrics/basic_parametric_estimation/__init__.py

@@ -0,0 +1,31 @@
+"""
+基础与参数估计模块
+提供普通最小二乘法 (OLS)、最大似然估计 (MLE) 和广义矩估计 (GMM) 方法
+"""
+
+# OLS模块
+from .ols import (
+    OLSResult,
+    ols_regression
+)
+
+# MLE模块
+from .mle import (
+    MLEResult,
+    mle_estimation
+)
+
+# GMM模块
+from .gmm import (
+    GMMResult,
+    gmm_estimation
+)
+
+__all__ = [
+    "OLSResult",
+    "ols_regression",
+    "MLEResult",
+    "mle_estimation",
+    "GMMResult",
+    "gmm_estimation"
+]

econometrics/basic_parametric_estimation/gmm/__init__.py

@@ -0,0 +1,13 @@
+"""
+广义矩估计 (GMM) 模块
+"""
+
+from .gmm_model import (
+    GMMResult,
+    gmm_estimation
+)
+
+__all__ = [
+    "GMMResult",
+    "gmm_estimation"
+]

econometrics/basic_parametric_estimation/gmm/gmm_model.py

@@ -0,0 +1,256 @@
+"""
+广义矩估计 (GMM) 模型实现
+"""
+
+from typing import List, Optional
+from pydantic import BaseModel, Field
+import numpy as np
+from scipy import stats
+
+
+class GMMResult(BaseModel):
+    """广义矩估计结果"""
+    coefficients: List[float] = Field(..., description="估计系数")
+    std_errors: List[float] = Field(..., description="系数标准误")
+    t_values: List[float] = Field(..., description="t统计量")
+    p_values: List[float] = Field(..., description="p值")
+    conf_int_lower: List[float] = Field(..., description="置信区间下界")
+    conf_int_upper: List[float] = Field(..., description="置信区间上界")
+    j_statistic: float = Field(..., description="J统计量")
+    j_p_value: float = Field(..., description="J统计量p值")
+    weight_matrix: List[List[float]] = Field(..., description="权重矩阵")
+    n_obs: int = Field(..., description="观测数量")
+    n_moments: int = Field(..., description="矩条件数量")
+    feature_names: List[str] = Field(..., description="特征名称")
+
+
+def _safe_inverse(matrix, reg_param=1e-10):
+    """安全的矩阵求逆函数"""
+    try:
+        # 检查矩阵是否为空或非二维
+        if matrix.size == 0 or matrix.ndim != 2:
+            raise ValueError("矩阵为空或不是二维数组")
+        
+        # 检查矩阵是否包含无效值
+        if np.isnan(matrix).any() or np.isinf(matrix).any():
+            raise ValueError("矩阵包含NaN或无穷大值")
+            
+        # 尝试直接求逆
+        return np.linalg.inv(matrix)
+    except np.linalg.LinAlgError:
+        # 如果矩阵奇异，添加正则化项
+        try:
+            # 确保矩阵是方阵
+            if matrix.shape[0] != matrix.shape[1]:
+                raise ValueError("矩阵不是方阵，无法求逆")
+                
+            # 添加正则化项
+            reg_matrix = matrix + np.eye(matrix.shape[0]) * reg_param
+            return np.linalg.inv(reg_matrix)
+        except np.linalg.LinAlgError:
+            # 如果仍然失败，使用伪逆
+            return np.linalg.pinv(matrix)
+
+
+def gmm_estimation(
+    y_data: List[float],
+    x_data: List[List[float]], 
+    instruments: Optional[List[List[float]]] = None,
+    feature_names: Optional[List[str]] = None,
+    constant: bool = True,
+    confidence_level: float = 0.95
+) -> GMMResult:
+    """
+    广义矩估计
+    
+    Args:
+        y_data: 因变量数据
+        x_data: 自变量数据
+        instruments: 工具变量数据 (如果为None，则使用x_data作为工具变量，退化为OLS)
+        feature_names: 特征名称
+        constant: 是否包含常数项
+        confidence_level: 置信水平
+        
+    Returns:
+        GMMResult: 广义矩估计结果
+        
+    Raises:
+        ValueError: 当输入数据无效时抛出异常
+    """
+    # 输入验证
+    if not y_data or not x_data:
+        raise ValueError("因变量和自变量数据不能为空")
+    
+    # 转换为numpy数组
+    y = np.array(y_data)
+    
+    # 确保X是二维数组
+    if isinstance(x_data[0], (int, float)):
+        # 单个特征的情况
+        X = np.array(x_data).reshape(-1, 1)
+    else:
+        X = np.array(x_data)
+    
+    # 验证数据维度一致性
+    if len(y) != X.shape[0]:
+        raise ValueError(f"因变量长度({len(y)})与自变量长度({X.shape[0]})不一致")
+    
+    n, k = X.shape
+    
+    # 处理工具变量
+    if instruments is None:
+        # 如果没有提供工具变量，则使用自变量作为工具变量（退化为OLS）
+        Z = X.copy()
+    else:
+        # 确保工具变量是二维数组
+        if isinstance(instruments[0], (int, float)):
+            Z = np.array(instruments).reshape(-1, 1)
+        else:
+            Z = np.array(instruments)
+        
+        # 验证工具变量维度
+        if len(Z) != len(y):
+            raise ValueError(f"工具变量长度({len(Z)})与因变量长度({len(y)})不一致")
+    
+    # 添加常数项
+    if constant:
+        X = np.column_stack([np.ones(n), X])
+        Z = np.column_stack([np.ones(n), Z])
+        if feature_names:
+            feature_names = ["const"] + feature_names
+        else:
+            feature_names = [f"const"] + [f"x{i}" for i in range(X.shape[1]-1)]
+    else:
+        if not feature_names:
+            feature_names = [f"x{i}" for i in range(X.shape[1])]
+    
+    # 手动实现GMM估计
+    try:
+        # 初始化权重矩阵为单位矩阵
+        W = np.eye(Z.shape[1])
+        
+        # 迭代估计直到收敛
+        for iteration in range(100):  # 最大迭代次数
+            # 一步GMM估计
+            # X'Z W Z'X beta = X'Z W Z'y
+            XZ = X.T @ Z
+            ZY = Z.T @ y
+            
+            # 更稳定的矩阵运算
+            left_side = XZ @ W @ XZ.T
+            right_side = XZ @ W @ ZY
+            
+            # 解线性方程组
+            try:
+                beta = np.linalg.solve(left_side, right_side)
+            except np.linalg.LinAlgError:
+                # 如果矩阵奇异，使用伪逆
+                beta = np.linalg.pinv(left_side) @ right_side
+            
+            # 计算残差
+            residuals = y - X @ beta
+            
+            # 更新权重矩阵（基于残差的矩条件）
+            moments = Z * residuals.reshape(-1, 1)
+            S = moments.T @ moments / n  # 协方差矩阵
+            
+            # 在更新权重矩阵前进行有效性检查
+            if np.isnan(S).any() or np.isinf(S).any():
+                raise ValueError("矩条件协方差矩阵包含无效值")
+            
+            # 安全地更新权重矩阵
+            W_new = _safe_inverse(S, reg_param=1e-8)  # 增加正则化参数以提高稳定性
+            
+            # 检查新权重矩阵的有效性
+            if np.isnan(W_new).any() or np.isinf(W_new).any():
+                raise ValueError("计算出的权重矩阵包含无效值")
+                
+            # 检查收敛性
+            if np.allclose(W, W_new, rtol=1e-6, atol=1e-10):
+                W = W_new
+                break
+            W = W_new
+        
+        # 计算最终的协方差矩阵和统计量
+        residuals = y - X @ beta
+        moments = Z * residuals.reshape(-1, 1)
+        S = moments.T @ moments / n
+        
+        # 检查矩条件协方差矩阵
+        if np.isnan(S).any() or np.isinf(S).any() or np.linalg.norm(S) == 0:
+            raise ValueError("矩条件协方差矩阵无效")
+        
+        # 计算系数协方差矩阵
+        # Var(beta) = (X'Z W Z'X)^(-1) X'Z W S W Z'X (X'Z W Z'X)^(-1)
+        XZ = X.T @ Z
+        
+        # 检查XZ矩阵
+        if np.isnan(XZ).any() or np.isinf(XZ).any():
+            raise ValueError("X'Z矩阵包含无效值")
+            
+        left_side = XZ @ W @ XZ.T
+        
+        # 检查左侧矩阵
+        if np.isnan(left_side).any() or np.isinf(left_side).any():
+            raise ValueError("左侧矩阵(X'Z W Z'X)包含无效值")
+        
+        left_side_inv = _safe_inverse(left_side, reg_param=1e-8)  # 使用相同正则化参数
+        
+        # 检查逆矩阵
+        if np.isnan(left_side_inv).any() or np.isinf(left_side_inv).any():
+            raise ValueError("左侧矩阵的逆包含无效值")
+        
+        # 计算协方差矩阵
+        cov_intermediate = XZ @ W @ S @ W @ XZ.T
+        if np.isnan(cov_intermediate).any() or np.isinf(cov_intermediate).any():
+            raise ValueError("中间协方差计算包含无效值")
+            
+        cov_beta = left_side_inv @ cov_intermediate @ left_side_inv
+        std_errors = np.sqrt(np.diag(cov_beta))
+        
+        # 避免零标准误
+        std_errors = np.maximum(std_errors, 1e-12)
+        
+        # 检查标准误
+        if np.isnan(std_errors).any() or np.isinf(std_errors).any():
+            raise ValueError("计算出的标准误包含无效值")
+        
+        # 计算t统计量和p值
+        t_values = beta / std_errors
+        p_values = 2 * (1 - stats.t.cdf(np.abs(t_values), n - len(beta)))
+        
+        # 计算置信区间
+        alpha = 1 - confidence_level
+        t_critical = stats.t.ppf(1 - alpha/2, n - len(beta))
+        conf_int_lower = beta - t_critical * std_errors
+        conf_int_upper = beta + t_critical * std_errors
+        
+        # J统计量（过度识别约束检验）
+        if Z.shape[1] > len(beta):
+            # 过度识别情况
+            moment_conditions = Z.T @ residuals
+            j_statistic = n * moment_conditions.T @ W @ moment_conditions
+            j_df = Z.shape[1] - len(beta)
+            j_p_value = 1 - stats.chi2.cdf(j_statistic, j_df)
+        else:
+            # 恰好识别情况
+            j_statistic = 0.0
+            j_p_value = 1.0
+        
+        return GMMResult(
+            coefficients=beta.tolist(),
+            std_errors=std_errors.tolist(),
+            t_values=t_values.tolist(),
+            p_values=p_values.tolist(),
+            conf_int_lower=conf_int_lower.tolist(),
+            conf_int_upper=conf_int_upper.tolist(),
+            j_statistic=float(j_statistic),
+            j_p_value=float(j_p_value),
+            weight_matrix=W.tolist(),
+            n_obs=n,
+            n_moments=Z.shape[1],
+            feature_names=feature_names
+        )
+    except Exception as e:
+        # 如果GMM失败，抛出异常
+        raise ValueError(f"GMM估计失败: {str(e)}")