代写论文：PM2.5对北京天气的影响
由于PM2.5在大气中的浓度依赖于前一时间段的浓度，故采用自回归方程建立上述方程。它需要一定的时间才能被排除在大气层之外。利用增广Dickey-Fuller检验对方程进行平稳性检验，发现残差是平稳的。采用breuss – godfrey LM检验进行自相关检验，无自相关。采用方差膨胀因子(VIF)检验多共线性，所有VIF均小于或接近5。因此，在上述估计中，多重共线性也不是一个问题。然而，最初的模型在使用breuss – pagan测试时被证明是异质塑性的。为了对其进行校正，模型中使用了稳健的标准误差。这些测试和初始模型显示在附录中。
代写论文：PM2.5对北京天气的影响
模型的R2为0.9153，说明模型解释了数据变化的91.53%。除风速(iws)在10%显著外，其余各变量在1%显著水平下均有统计学意义。
代写论文：PM2.5对北京天气的影响
The above equation is developed using an autoregressive equation because the concentration of PM2.5 in the atmosphere is dependent on the concentration in the previous time period. It takes a certain time for it to be excluded from the atmosphere. The equation was tested for stationarity using Augmented Dickey-Fuller test and the residuals were found to be stationary. Autocorrelation was tested for by Breusch-Godfrey LM test which showed that there was no autocorrelation. Variance Inflation Factor (VIF) was used to test for multicollinearity and all the VIFs were either less than or close to 5. Therefore, it was concluded that multicollinearity is not a problem in the above estimation either. However, the initial model turned out to be heteroskedastic when tested for it using Breusch-Pagan test. To correct for it, robust standard errors are used in the model. These tests and the initial model are shown in the Appendix.
代写论文：PM2.5对北京天气的影响
The R2 of the model is 0.9153 which implies that the model explains 91.53% of the variation in the data. All the variables are statistically significant at 1% level of significance except for wind speed (iws) which is significant at 10% level of significance.