Question

我试图绘制经典＆＃34; Fitted vs Residual＆＃34;时间序列线性模型对fpp包中花式时间序列的绘图：

结构（c（1664.81,2397.53,2840.71,3547.29,3752.96,3714.74， 4349.61,3566.34,5021.82,6423.48,7600.6,19756.21,2499.81， 5198.24,7225.14,4806.03,59.00.88,4951.34,6179.12,4752.15， 5496.43,5835.1,12600.08,28541.72,4717.02,5702.63,9957.58， 5304.78,6492.43,6630.8,7349.62,8176.62,8573.17,9690.5， 15151.84,34061.01,5921.1,5814.58,12421.25,6369.77,7609.12， 7224.75,8121.22,7799.25,8093.06,8476.7,17914.66,30114.41， 4826.64,6470.23,9638.77,8821.17,8722.37,10209.48,11276.55， 12552.22,11637.39,13606.89,21822.11,45060.69,7615.03,9849.69， 14558.4,11587.33,9332.56,13082.09,16732.78,1888.61,23933.38， 25391.35,36024.8,80721.71,10243.24,11266.88,21882.84,17357.33， 15997.79,18601.53,26155.15,28586.52,30505.41,30821.33,46634.38， 104660.67）,. Tsp = c（1987,1993.91666666667,12），class =＆＃34; ts＆＃34;）

library(fpp)
fit = tslm(fancy ~ trend + season)
plot(fitted(fit), residuals(fit), xlab = "Predicted scores", ylab = "Residuals")

情节很混乱，因为拟合（拟合）和残差（拟合）又是月度时间序列对象，因此散点图不起作用。

如何在正常的lm中像往常一样显示散点图？

感谢您的帮助。

Answer 1

谢谢大家，

我通过在绘图之前将ts转换为向量找到了一个快速的转变：

fit_vector <- as.vector(fitted(fit))
fit_residuals <- as.vector(residuals(fit))
plot(fit_vector, fit_residuals, xlab = "Predicted scores", ylab = "Residuals")

在月度时间序列线性模型中拟合Vs残差

1 个答案: