为何R语言中随机森林训练集的拟合优度与手工计算结果不一致？-经管之家官网！

请教各位大佬一个问题，在使用R语言randomForestSRC包进行随机森林回归预测时，训练集的拟合优度为35.3%，但是通过predict函数输出训练集的预测值进行手工计算时R方变成了90.3%。个人猜测是因为训练集的数据已经被模型熟悉，那如何才能输出R方为35.3%时的预测值呢？因为看到很多文献都有训练集与验证集的精度对比。（下面是我的代码，求大佬们指点一二！）万分感谢！> traindataend <- rfsrc(容重 ~ ., data = traindata, mtry=5, nodesize=1, ntree = 1000)#最终模型
> print(traindataend)#输出拟合优度
Sample size: 561
Number of trees: 1000
Forest terminal node size: 1
Average no. of terminal nodes: 315.438
No. of variables tried at each split: 5
Total no. of variables: 7
Resampling used to grow trees: swor
Resample size used to grow trees: 355
Analysis: RF-R
Family: regr
Splitting rule: mse *random*
Number of random split points: 10
% variance explained: 35.3
Error rate: 0.01
traindataPREDICT <- predict(traindataend, newdata = traindata)#对训练集进行预测