计量经济学是数学、统计技术和经济分析的综合,即运用数学、统计方法和相关经济理论,通过计量模型来揭示经济数量关系和规律。R语言包,已经实现了现代计量经济学的很多统计分析功能,下面从面板数据模型和时间序列模型简要介绍R语言相关程序包:
1.面板数据模型,Panel data models,23个包
2.时间序列模型,Time series data and models,19个包
stats包:一般的方法是用stats包中函数,如 lm()或者glm(),对混合截面模型进行估计。
sandwich包:提供了稳健性标准误的估算
https://cloud.r-project.org/web/packages/sandwich/index.html
clusterSEs包:计算聚类稳健的p值和置信区间
https://cloud.r-project.org/web/packages/clusterSEs/index.html
pcse包:面板校正标准误估计
https://cloud.r-project.org/web/packages/pcse/index.html
clubSandwich包:最小样本校正的稳健性聚类方差估计
https://cloud.r-project.org/web/packages/clubSandwich/index.html
plm包:面板数据线性模型,包括动态面板模型
https://cloud.r-project.org/web/packages/plm/index.html
Paneldata包:固定效应模型和随机效应模型
https://cloud.r-project.org/web/packages/Paneldata/index.html
OrthoPanels包:固定效应动态面板模型
https://cloud.r-project.org/web/packages/OrthoPanels/index.html
feisr包:提供固定效应个体斜率模型(fixed effects individual slope models)
https://cloud.r-project.org/web/packages/feisr/index.html
panelr包:提供中间(或混合)面板模型,包括多层、GEE和贝叶斯等模型
https://cloud.r-project.org/web/packages/panelr/index.html
panelvar包:面板向量自回归
https://cloud.r-project.org/web/packages/panelvar/index.html
pglm包:提供了面板数据GLM-like模型估计
https://cloud.r-project.org/web/packages/pglm/index.html
geepack包:广义估计方程包,如GEE模型
https://cloud.r-project.org/web/packages/geepack/index.html
lme4包:线性混合效应模型
https://cloud.r-project.org/web/packages/lme4/index.html
nlme包:线性和非线性混合效应模型
https://cloud.r-project.org/web/packages/nlme/index.html
ivfixed包:工具变量固定效应面板模型
https://cloud.r-project.org/web/packages/ivfixed/index.html
ivpanel包:工具变量面板模型,适合固定效应、随机效应和两者
https://cloud.r-project.org/web/packages/ivpanel/index.html
phtt包:当未观察到的异质性影响随时间变化时,提供了分析具有较大维度n和T的面板数据的可能性
https://cloud.r-project.org/web/packages/phtt/index.html
wahc包:固定效应面板模型中的自相关和异方差校正
https://cloud.r-project.org/web/packages/wahc/index.html
panelAR包:具有横断面异方差或相关性的线性AR(1)面板模型的估计
https://cloud.r-project.org/web/packages/panelAR/index.html
PANICr包:非平稳性的PANIC测试
https://cloud.r-project.org/web/packages/PANICr/index.html
pdR包:面板数据中的单位根检验
https://cloud.r-project.org/web/packages/pdR/index.html
pampe包:用于方案评估的面板数据方法
https://cloud.r-project.org/web/packages/pampe/index.html
stats包:封装在stats包中的“ts”对规则间隔时间序列的设置,用于年、季和月度等数据
zoo包:封装在zoo包中的"zooreg",构造规则时间序列对象
https://cloud.r-project.org/web/packages/zoo/index.html
zoo包:提供规则和不规则间隔时间序列的基础结构
https://cloud.r-project.org/web/packages/zoo/index.html
xts包:可扩展时间序列,基于时间数据进行统一处理
https://cloud.r-project.org/web/packages/xts/index.html
stats包:使用ar()和ARIMA建模来估算简单的自回归模型,使用arima()进行Box-Jenkins类型分析
forecast包:时间序列和线性模型的预测功能
https://cloud.r-project.org/web/packages/forecast/index.html
dynlm包:动态线性回归
https://cloud.r-project.org/web/packages/dynlm/index.html
nlme包:使用nlme的gls()通过GLS进行带有AR误差项的线性回归模型
https://cloud.r-project.org/web/packages/nlme/index.html
stats包:标准模型可用stats包中StructTS()函数
stats包:decompose() 和 HoltWinters()
stats包:简单模型用stats包中的 ar()
vars 包:提供更详细的模型,包括回归诊断、可视化功能
https://cloud.r-project.org/web/packages/vars/index.html
panelvar包:面板数据中向量自回归
https://cloud.r-project.org/web/packages/panelvar/index.html
urca包:时间序列数据的单位根和协整检验
https://cloud.r-project.org/web/packages/urca/index.html
tseries包:时间序列分析与计算
https://cloud.r-project.org/web/packages/tseries/index.html
CADFtest包:协变量Dickey-Fuller单位根检验
https://cloud.r-project.org/web/packages/CADFtest/index.html
pco包:面板协整检验
https://cloud.r-project.org/web/packages/pco/index.html
tsDyn包:阈值和平滑转换模型
https://cloud.r-project.org/web/packages/tsDyn/index.html
PSTR 包:面板平滑转换回归模型
https://cloud.r-project.org/web/packages/PSTR/index.html
midasr包:MIDAS回归和其他混合频率时间序列数据分析
https://cloud.r-project.org/web/packages/midasr/index.html
gets包:General-to-Specific (GETS)模型和指标饱和度方法
https://cloud.r-project.org/web/packages/gets/index.html
tsfa包:时间序列因子分析
https://cloud.r-project.org/web/packages/tsfa/index.html
bimets包:联立方程模型对时间序列数据进行计量经济学建模
https://cloud.r-project.org/web/packages/bimets/index.html
dlsem包:分布滞后线性结构方程模型
https://cloud.r-project.org/web/packages/dlsem/index.html
apt包:非对称价格传递模型
https://cloud.r-project.org/web/packages/apt/index.html
注:文中所列出R语言包,均参考CRAN Task Views官网信息,获取全文详细内容,请点击阅读原文进行查阅!
文章首发于公众号数据Seminar
工具&方法 | “名牌包”:面板、时间序列模型常用R语言包mp.weixin.qq.com更多精彩内容请关注公众号数据Seminar