Question

我根据以下建议的解决方案修改了编辑内容：

library(caret)
library(ggplot2)
library('scales')


Macro <- read.csv("P:/R/Macro.csv")

foo <- function (start) {

  mycontrol <- trainControl(method = "timeslice",
                            initialWindow = 8,
                            horizon = 1,
                            fixedWindow = FALSE, 
                            savePredictions = TRUE)


  myfit <- train(Y ~ X1372C + S1 + S2 + S3,
                 data = Macro[start:14, ,drop = FALSE],
                 method = "lm",
                 trControl = mycontrol)

  c(myfit$pred)  ## return; drop dimension as a vector
}


Forecast <- sapply(1:6, foo)  ## will simplify to matrix

dput(Forecast)

这转化为此输出：

警告讯息： 1：在nominalTrainWorkflow中（x = x，y = y，wts = weights，info = trainInfo，：重新采样的绩效指标中存在缺失值。 2：在nominalTrainWorkflow中（x = x，y = y，wts = weights，info = trainInfo，：重新采样的绩效指标中存在缺失值。 3：在nominalTrainWorkflow中（x = x，y = y，wts = weights，info = trainInfo，：重新采样的绩效指标中存在缺失值。 4：在nominalTrainWorkflow（x = x，y = y，wts = weights，info = trainInfo，：重新采样的绩效指标中存在缺失值。 5：在nominalTrainWorkflow中（x = x，y = y，wts = weights，info = trainInfo，：重新采样的绩效指标中存在缺失值。 6：在nominalTrainWorkflow（x = x，y = y，wts = weights，info = trainInfo，：重新抽样的绩效指标中缺少值。

dput（预测）结构（c（921.172565647808,926.099130620101,1240.09797589312， 942.482635914736,955.942081354932,977.408884588125），c（977.0702941， 943.8941538,1272.797631,981.859451,995.574098,1052.287953 ），9：14，c（TRUE，TRUE，TRUE，TRUE，TRUE，TRUE），c（＆＃34; Training1＆＃34;，＆＃34;培训2＆＃34;，＆＃34;培训3＆＃34;，＆＃34;培训4＆＃34;，＆＃34;培训5＆＃34;，＆＃34;培训6＆＃34; ），c（926.678510620924,1241.54853541131,943.689699182586,952.297326382802， 979.924409311149），c（943.8941538,1272.797631,981.859451,995.574098， 1052.287953），9：13，c（TRUE，TRUE，TRUE，TRUE，TRUE），c（＆＃34; Training1＆＃34;，＆＃34;培训2＆＃34;，＆＃34;培训3＆＃34;，＆＃34;培训4＆＃34;，＆＃34;培训5＆＃34;），c（1257.33305898499， 951.366992832791,957.163715037822,986.364307731912），c（1272.797631， 981.859451,995.574098,1052.287953），9：12，c（TRUE，TRUE，TRUE， TRUE），c（＆＃34; Training1＆＃34;，＆＃34; Training2＆＃34;，＆＃34; Training3＆＃34;，＆＃34; Training4＆＃34;）， c（993.427314382052,978.520174690649,1067.16914028792）， c（981.859451,995.574098,1052.287953），9：11，c（TRUE，TRUE， TRUE），c（＆＃34; Training1＆＃34;，＆＃34; Training2＆＃34;，＆＃34; Training3＆＃34;），c（987.694506364155， 1102.00413769988），c（995.574098,1052.287953），9：10，c（TRUE， TRUE），c（＆＃34; Training1＆＃34;，＆＃34; Training2＆＃34;），1058.30501548728,1052.287953， 9L，TRUE，＆＃34; Training1＆＃34;）,. Dim = 5：6，.Dimnames = list（c（＆＃34; pred＆＃34;，＆＃34; obs＆＃34;，＆＃34; rowIndex＆＃34;，＆＃34; intercept＆＃34;，＆＃34; Resample＆＃34;），NULL））

dput（Macro [1:14，]）结构（列表（Qtrs =结构（1:14，.Label = c（＆＃34; 14_Q2＆＃34;，＆＃34; 14_Q3＆＃34;，＆＃34; 14_Q4＆＃34;，＆＃34; 15_Q1＆＃34;，＆＃34; 15_Q2＆＃34;，＆＃34; 15_Q3＆＃34;，＆＃34; 15_Q4＆＃34;，＆＃34; 16_Q1＆＃34;，＆＃34; 16_Q2＆＃34;，＆＃34; 16_Q3＆＃34;，＆＃34; 16_Q4＆＃34;，＆＃34; 17_Q1＆＃34;，＆＃34; 17_Q2＆＃34;，＆＃34; 17_Q3＆＃34;），class =＆＃ 34;因子＆＃34）， Y = c（865.8283846,856.0688462,1165.958691,883.3432512， 902.1949231,897.0950769,1237.512923,962.5309231,977.0702941， 943.8941538,1272.797631,981.859451,995.574098,1052.287953 ），X1372C = c（0.906148417,0.882519787,1.329575232,1.088817819， 1.180987191,1.167632493,1.593644333,1.178410516,1.2626391569， 1.221516083,1.605735851,1.202507143,1.243837663,1.359676971 ），X5244C = c（0.906148417,0.882519787,1.329575232,1.088817819， 1.180987191,1.167632493,1.593644333,1.178410516,1.2626391569， 1.221516083,1.605735851,1.202507143,1.243837663,1.359676971 ），X5640C = c（0.668605303,0.650010329,0.871325692,0.657860513， 0.717979329,0.734934807,0.987398267,0.729793323,0.811565803， 0.820667752,1.072773575,0.783936578,0.864640587,0.955630807 ），X6164C = c（0.405020101,0.372197584,0.518686161,0.373651444， 0.407590328,0.406092873,0.560442245,0.417412494,0.429938649， 0.461566266,0.592980109,0.444907445,0.47447048,0.558013448 ），X1372B = c（1.874539814,1.816307314,2.538860638,1.785564496， 1.877335003,1.896950271,2.677640702,2.057301773,2.060862256， 2.010340805,2.749691737,2.112415534,2.123956718,2.264930431 ），X5244B = c（1.874539814,1.816307314,2.538860638,1.785564496， 1.877335003,1.896950271,2.677640702,2.057301773,2.060862256， 2.010340805,2.749691737,2.112415534,2.123956718,2.264930431 ），X5640B = c（2.784147985,2.728887944,3.915195977,3.036215525， 3.037519069,3.035750269,4.235355932,3.304887671,3.26268204， 3.141199768,4.294197969,3.274759241,3.2111118188,3.501393693 ），X6164B = c（0.791458844,0.78168066,1.18611112,0.859852789， 0.944191582,0.976372468,1.397898629,1.099556909,1.128398283， 1.131976904,1.580815657,1.206596743,1.245306897,1.358211802 ），X1372V = c（0.097143401,0.097524773,0.141261151,0.115168338， 0.120877067,0.119050862,0.157452868,0.117294253,0.11874258， 0.117416967,0.154157986,0.117553225,0.117734259,0.126540804 ），X5244V = c（0.097143401,0.097524773,0.141261151,0.115168338， 0.120877067,0.119050862,0.157452868,0.117294253,0.11874258， 0.117416967,0.154157986,0.117553225,0.117734259,0.126540804 ），X5640V = c（0.064650539,0.063541823,0.083794768,0.063561861， 0.06730385,0.068859437,0.089436159,0.066723151,0.072244789， 0.072641489,0.095058099,0.070520087,0.075122879,0.082610429 ），X6164V = c（0.044563645,0.042226354,0.057467132,0.042275768， 0.043796282,0.043972939,0.059422138,0.044436698,0.044714835， 0.047485033,0.060888955,0.046498904,0.047787208,0.055743044 ），X1372BV = c（0.211443705,0.20868086,0.281834069,0.198443943， 0.205496416,0.207598984,0.285816341,0.21745711,0.21640918， 0.210584615,0.283752378,0.218344658,0.215849274,0.229469246 ），X5244BV = c（0.211443705,0.20868086,0.281834069,0.198443943， 0.205496416,0.207598984,0.285816341,0.21745711,0.21640918， 0.210584615,0.283752378,0.218344658,0.215849274,0.229469246 ），X5640BV = c（0.27897666,0.2771921,0.387651075,0.297277243， 0.293774368,0.293731964,0.400189046,0.309076866,0.302470138， 0.29076674,0.39564161,0.301178985,0.297513626,0.315261748 ），X6164BV = c（0.087276144,0.089506522,0.130605096,0.094950656， 0.101815228,0.105435663,0.148800072,0.115872327,0.118331193， 0.118122767,0.163058,0.124425055,0.126056291,0.136434954 ），S1 = c（1L，0L，0L，0L，1L，0L，0L，0L，1L，0L，0L，0L， 1L，0L），S2 = c（0L，1L，0L，0L，0L，1L，0L，0L，0L，1L，0L， 0L，0L，1L），S3 = c（0L，0L，1L，0L，0L，0L，1L，0L，0L，0L， 1L，0L，0L，0L））,. Name = c（＆＃34; Qtrs＆＃34;，＆＃34; Y＆＃34;，＆＃34; X1372C＆＃34;，＆＃34; X5244C＆＃34; ，＆＃34; X5640C＆＃34;，＆＃34; X6164C＆＃34;，＆＃34; X1372B＆＃34;，＆＃34; X5244B＆＃34;，＆＃34; X5640B＆＃34;，＆＃34; X6164B＆＃34;，＆＃34; X1372V＆＃34;，＆＃34; X5244V＆＃34;，＆＃34; X5640V＆＃34;，＆＃34; X6164V＆＃34;，＆＃34; X1372BV＆＃34;，＆＃34; X5244BV＆＃34;，＆＃34; X5640BV＆＃34 ;, ＆＃34; X6164BV＆＃34;，＆＃34; S1＆＃34;，＆＃34; S2＆＃34;，＆＃34; S3＆＃34;），row.names = c（NA，14L），class = ＆＃34; data.frame＆＃34）

Answer 1

您的条款仅在使用的data中有所不同。更准确地说，是数据集Macro的子集。更准确地说，是子集的起始行。那么为什么不写一个函数呢？

foo <- function (start) {

  mycontrol <- trainControl(method = "timeslice",
                            initialWindow = 8,
                            horizon = 1,
                            fixedWindow = FALSE, 
                            savePredictions = TRUE)


  myfit <- train(Y ~ X1372C + S1 + S2 + S3,
                 data = Macro[start:14, ,drop = FALSE],
                 method = "lm",
                 trControl = mycontrol)

  c(myfit$pred)  ## return; drop dimension as a vector
  }

然后

Forecast <- sapply(1:6, foo)  ## will simplify to matrix

最后，您的F9到F14只是矩阵Forecast的下三角（加上对角线）的行和：

upper.tri(Forecast) <- 0
rowSums(Forecast)

Answer 2

我很快就完成了你的代码，但这应该可行。我建议您查看defining functions in R。

clauseLoop <- function(startPoint) {
    myControl <- trainControl(method = "timeslice",
                      initialWindow = 8,
                      horizon = 1,
                      fixedWindow = FALSE, 
                      savePredictions = TRUE)

    myFit <- train(Y ~ X1372C + S1 + S2 + S3,
                   data = startPoint,
                   method = "lm",
                   trControl = mycontrol)

    Forecast <- myFit$pred
}

ForecastX <- clauseLoop("x <- Macro[6:14,]")

如何避免6次写同一个R子句

2 个答案: