使用大型数据集的 For 循环的极端处理时间
Extreme Processing Time on For Loop Using Large Dataset
我准备了一个例程，将一个单元格（“DBH”）的值复制到另一行中的另一个单元格（“DG”），其中被复制到的行紧接在被复制的行之后（及时，不一定在table）。
PlotID 'R8' 表示第 8 个测量值，'R7' 第 7 个，依此类推。
该循环旨在检查该行是否具有先前的测量值，如果有，则记录该测量值的唯一行标识符。然后它使用前面的行标识符来搜索要通过第二个嵌套循环复制的前面的记录。
问题是我使用的完整数据集包含超过 650,000 条记录。我知道这种结构会有些低效，但是当我 运行 它时 运行 时间每小时仅移动大约 10,000 条记录。有什么提高计算效率的建议吗？
# sort TreeInit Table to facilitate desired iteration
FVS_TreeInit$StandPlotTree_ID <- paste(FVS_TreeInit$StandPlot_ID,"T",FVS_TreeInit$Tree_ID,sep="")
# loop through FVS_TreeInit and add previous diameter increment identifier to current row
for (i in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[i,"Plot_ID"] == "R8"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R7T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R7"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R6T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R6"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R5T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R5"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R4T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R4"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R3T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R3"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R2T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R2"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R1T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R1"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else{
FVS_TreeInit[i,"GD"] <- NA
}
if (i%%1000 == 0){
print(paste(i," rows iterated...",sep=""))
}
}
}
一些样本数据（.csv）：
Stand_ID,Plot_ID,StandPlot_ID,Tree_ID,DBH,DG,Species,Age,StandPlotTree_ID,StandPlotTree_ID_Past
D18P186001,R0,D18P186001R0,1,4.76378225,,ABBA,,D18P186001R0T1,
D18P186001,R0,D18P186001R0,2,1.259843219,,BEPA,,D18P186001R0T2,
D18P186001,R0,D18P186001R0,3,0.275590695,,BEPA,,D18P186001R0T3,
D18P186001,R0,D18P186001R0,4,0.629921609,,BEPA,,D18P186001R0T4,
D18P186001,R0,D18P186001R0,5,0.1968505,,BEPA,,D18P186001R0T5,
D18P186001,R0,D18P186001R0,6,3.70078925,,ABBA,26,D18P186001R0T6,
D18P186001,R0,D18P186001R0,7,2.834647125,,ABBA,,D18P186001R0T7,
D18P186001,R0,D18P186001R0,8,1.259843219,,,,D18P186001R0T8,
D18P186001,R0,D18P186001R0,9,2.283465875,,,,D18P186001R0T9,
D18P186001,R0,D18P186001R0,10,1.181103,,,,D18P186001R0T10,
D18P186001,R0,D18P186001R0,11,1.535433938,,BEPA,,D18P186001R0T11,
D18P186001,R0,D18P186001R0,12,4.212600625,,ABBA,,D18P186001R0T12,
D18P186001,R0,D18P186001R0,13,3.30708825,,ABBA,,D18P186001R0T13,
D18P186001,R0,D18P186001R0,14,1.259843219,,ABBA,,D18P186001R0T14,
D18P186001,R0,D18P186001R0,15,4.212600625,,ABBA,28,D18P186001R0T15,
D18P186001,R0,D18P186001R0,16,3.149608,,ABBA,28,D18P186001R0T16,
D18P186001,R0,D18P186001R0,17,1.181103,,ABBA,,D18P186001R0T17,
D18P186001,R0,D18P186001R0,18,0.944882438,,BEPA,,D18P186001R0T18,
D18P186001,R0,D18P186001R0,19,0.393701,,BEPA,,D18P186001R0T19,
D18P186001,R0,D18P186001R0,20,0.314960805,,BEPA,,D18P186001R0T20,
D18P186001,R0,D18P186001R0,21,2.362206,,ABBA,22,D18P186001R0T21,
D18P186001,R0,D18P186001R0,22,0.314960805,,BEPA,,D18P186001R0T22,
D18P186001,R0,D18P186001R0,23,1.062992719,,ABBA,,D18P186001R0T23,
D18P186001,R0,D18P186001R0,24,3.93701,,ABBA,25,D18P186001R0T24,
D18P186001,R0,D18P186001R0,25,4.76378225,,ABBA,,D18P186001R0T25,
D18P186001,R0,D18P186001R0,26,0.748031891,,BEPA,,D18P186001R0T26,
D18P186001,R0,D18P186001R0,27,0.944882438,,BEPA,,D18P186001R0T27,
D18P186001,R0,D18P186001R0,28,3.267718375,,ABBA,,D18P186001R0T28,
D18P186001,R0,D18P186001R0,29,0.314960805,,BEPA,,D18P186001R0T29,
D18P186001,R0,D18P186001R0,30,0.472441219,,BEPA,,D18P186001R0T30,
D18P186001,R0,D18P186001R0,31,4.606301625,,ABBA,,D18P186001R0T31,
D18P186001,R0,D18P186001R0,32,4.055120375,,ABBA,,D18P186001R0T32,
D18P186001,R0,D18P186001R0,33,0.433071109,,,,D18P186001R0T33,
D18P186001,R0,D18P186001R0,34,1.023622562,,,,D18P186001R0T34,
D18P186001,R0,D18P186001R0,35,0.5905515,,,,D18P186001R0T35,
D18P186001,R0,D18P186001R0,36,0.748031891,,,,D18P186001R0T36,
D18P186001,R0,D18P186001R0,37,3.425198625,,ABBA,30,D18P186001R0T37,
D18P186001,R0,D18P186001R0,38,4.29134075,,ABBA,,D18P186001R0T38,
D18P186001,R0,D18P186001R0,39,3.93701,,ABBA,,D18P186001R0T39,
D18P186001,R0,D18P186001R0,40,4.251970875,,ABBA,,D18P186001R0T40,
D18P186001,R0,D18P186001R0,41,4.09449025,,ABBA,,D18P186001R0T41,
D18P186001,R0,D18P186001R0,42,6.181105625,,ABBA,24,D18P186001R0T42,
D18P186001,R0,D18P186001R0,43,3.858269875,,ABBA,,D18P186001R0T43,
D18P186001,R0,D18P186001R0,44,3.7401595,,ABBA,,D18P186001R0T44,
D18P186001,R0,D18P186001R0,45,0.5905515,,ABBA,,D18P186001R0T45,
D18P186001,R0,D18P186001R0,46,1.259843219,,ABBA,,D18P186001R0T46,
D18P186001,R0,D18P186001R0,47,3.89763975,,ABBA,,D18P186001R0T47,
D18P186001,R0,D18P186001R0,48,0.708661781,,ABBA,,D18P186001R0T48,
D18P186001,R0,D18P186001R0,49,0.314960805,,ABBA,,D18P186001R0T49,
D18P186001,R0,D18P186001R0,50,4.1338605,,ABBA,,D18P186001R0T50,
D18P186001,R0,D18P186001R0,51,1.889764875,,BEPA,,D18P186001R0T51,
D18P186001,R0,D18P186001R0,52,0.314960805,,,,D18P186001R0T52,
D18P186001,R0,D18P186001R0,53,0.629921609,,BEPA,,D18P186001R0T53,
D18P186001,R0,D18P186001R0,54,0.354330891,,BEPA,,D18P186001R0T54,
D18P186001,R0,D18P186001R0,55,1.456693719,,BEPA,,D18P186001R0T55,
D18P186001,R0,D18P186001R0,56,2.047245125,,,,D18P186001R0T56,
D18P186001,R0,D18P186001R0,57,2.480316375,,,,D18P186001R0T57,
D18P186001,R0,D18P186001R0,58,1.968505,,,,D18P186001R0T58,
D18P186001,R0,D18P186001R0,59,3.93701,,ABBA,,D18P186001R0T59,
D18P186001,R0,D18P186001R0,60,1.102362781,,ABBA,,D18P186001R0T60,
D18P186001,R0,D18P186001R0,61,0.472441219,,BEPA,,D18P186001R0T61,
D18P186001,R0,D18P186001R0,62,0.905512281,,BEPA,,D18P186001R0T62,
D18P186001,R0,D18P186001R0,63,0.944882438,,BEPA,,D18P186001R0T63,
D18P186001,R0,D18P186001R0,64,1.574804,,ABBA,,D18P186001R0T64,
D18P186001,R0,D18P186001R0,65,0.472441219,,BEPA,,D18P186001R0T65,
D18P186001,R0,D18P186001R0,66,5.15748325,,ABBA,,D18P186001R0T66,
D18P186001,R0,D18P186001R0,67,5.35433375,,ABBA,29,D18P186001R0T67,
D18P186001,R0,D18P186001R0,68,0.511811281,,ABBA,,D18P186001R0T68,
D18P186001,R0,D18P186001R0,69,4.68504175,,ABBA,,D18P186001R0T69,
D18P186001,R0,D18P186001R0,70,6.8897675,,ABBA,23,D18P186001R0T70,
D18P186001,R0,D18P186001R0,71,0.275590695,,BEPA,,D18P186001R0T71,
D18P186001,R0,D18P186001R0,72,4.606301625,,ABBA,20,D18P186001R0T72,
D18P186001,R0,D18P186001R0,73,0.511811281,,BEPA,,D18P186001R0T73,
D18P186001,R0,D18P186001R0,74,0.275590695,,BEPA,,D18P186001R0T74,
D18P186001,R0,D18P186001R0,75,0.393701,,BEPA,,D18P186001R0T75,
D18P186001,R0,D18P186001R0,76,0.472441219,,BEPA,,D18P186001R0T76,
D18P186001,R0,D18P186001R0,77,0.669291719,,ABBA,,D18P186001R0T77,
D18P186001,R0,D18P186001R0,78,0.354330891,,BEPA,,D18P186001R0T78,
D18P186001,R0,D18P186001R0,79,4.409451125,,ABBA,,D18P186001R0T79,
D18P186001,R0,D18P186001R0,80,7.12598825,,ABBA,,D18P186001R0T80,
D18P186001,R0,D18P186001R0,81,4.724412,,ABBA,,D18P186001R0T81,
D18P186001,R1,D18P186001R1,1,5.86614475,,ABBA,,D18P186001R1T1,
D18P186001,R1,D18P186001R1,2,1.3779535,,BEPA,,D18P186001R1T2,
D18P186001,R1,D18P186001R1,3,,,BEPA,,D18P186001R1T3,
D18P186001,R1,D18P186001R1,4,,,BEPA,,D18P186001R1T4,
D18P186001,R1,D18P186001R1,5,0.236220609,,BEPA,,D18P186001R1T5,
D18P186001,R1,D18P186001R1,6,4.1338605,,ABBA,31,D18P186001R1T6,
D18P186001,R1,D18P186001R1,7,3.425198625,,ABBA,,D18P186001R1T7,
D18P186001,R1,D18P186001R1,9,2.480316375,,,,D18P186001R1T9,
D18P186001,R1,D18P186001R1,10,1.259843219,,,,D18P186001R1T10,
D18P186001,R1,D18P186001R1,11,1.574804,,BEPA,,D18P186001R1T11,
D18P186001,R1,D18P186001R1,12,4.88189225,,ABBA,,D18P186001R1T12,
D18P186001,R1,D18P186001R1,13,3.89763975,,ABBA,,D18P186001R1T13,
D18P186001,R1,D18P186001R1,14,1.3779535,,ABBA,,D18P186001R1T14,
D18P186001,R1,D18P186001R1,15,5.196853125,,ABBA,33,D18P186001R1T15,
D18P186001,R1,D18P186001R1,16,4.055120375,,ABBA,33,D18P186001R1T16,
D18P186001,R1,D18P186001R1,17,1.220473062,,ABBA,,D18P186001R1T17,
D18P186001,R1,D18P186001R1,18,0.9842525,,BEPA,,D18P186001R1T18,
D18P186001,R1,D18P186001R1,19,0.393701,,BEPA,,D18P186001R1T19,
D18P186001,R1,D18P186001R1,20,0.472441219,,BEPA,,D18P186001R1T20,
D18P186001,R1,D18P186001R1,21,2.716536938,,ABBA,27,D18P186001R1T21,
D18P186001,R1,D18P186001R1,22,,,BEPA,,D18P186001R1T22,
D18P186001,R1,D18P186001R1,23,1.181103,,ABBA,,D18P186001R1T23,
D18P186001,R1,D18P186001R1,24,5.000002625,,ABBA,30,D18P186001R1T24,
D18P186001,R1,D18P186001R1,25,4.88189225,,ABBA,,D18P186001R1T25,
D18P186001,R1,D18P186001R1,26,,,BEPA,,D18P186001R1T26,
D18P186001,R1,D18P186001R1,27,1.023622562,,BEPA,,D18P186001R1T27,
D18P186001,R1,D18P186001R1,28,4.330711,,ABBA,,D18P186001R1T28,
D18P186001,R1,D18P186001R1,29,0.354330891,,BEPA,,D18P186001R1T29,
D18P186001,R1,D18P186001R1,30,0.511811281,,BEPA,,D18P186001R1T30,
D18P186001,R1,D18P186001R1,31,5.74803475,,ABBA,,D18P186001R1T31,
D18P186001,R1,D18P186001R1,32,5.27559325,,ABBA,,D18P186001R1T32,
D18P186001,R1,D18P186001R1,33,,,,,D18P186001R1T33,
D18P186001,R1,D18P186001R1,34,0.944882438,,,,D18P186001R1T34,
D18P186001,R1,D18P186001R1,35,0.669291719,,,,D18P186001R1T35,
D18P186001,R1,D18P186001R1,36,1.062992719,,,,D18P186001R1T36,
D18P186001,R1,D18P186001R1,37,4.5275615,,ABBA,35,D18P186001R1T37,
D18P186001,R1,D18P186001R1,38,5.629924375,,ABBA,,D18P186001R1T38,
D18P186001,R1,D18P186001R1,39,4.88189225,,ABBA,,D18P186001R1T39,
D18P186001,R1,D18P186001R1,40,5.47244375,,ABBA,,D18P186001R1T40,
D18P186001,R1,D18P186001R1,41,4.842522375,,ABBA,,D18P186001R1T41,
D18P186001,R1,D18P186001R1,42,7.204728,,ABBA,29,D18P186001R1T42,
D18P186001,R1,D18P186001R1,43,4.842522375,,ABBA,,D18P186001R1T43,
D18P186001,R1,D18P186001R1,44,4.606301625,,ABBA,,D18P186001R1T44,
D18P186001,R1,D18P186001R1,45,0.748031891,,ABBA,,D18P186001R1T45,
D18P186001,R1,D18P186001R1,46,1.3779535,,ABBA,,D18P186001R1T46,
D18P186001,R1,D18P186001R1,47,4.37008125,,ABBA,,D18P186001R1T47,
D18P186001,R1,D18P186001R1,48,0.9842525,,ABBA,,D18P186001R1T48,
D18P186001,R1,D18P186001R1,49,0.354330891,,ABBA,,D18P186001R1T49,
D18P186001,R1,D18P186001R1,50,5.07874275,,ABBA,,D18P186001R1T50,
D18P186001,R1,D18P186001R1,51,2.401576062,,BEPA,,D18P186001R1T51,
D18P186001,R1,D18P186001R1,52,,,,,D18P186001R1T52,
D18P186001,R1,D18P186001R1,53,,,BEPA,,D18P186001R1T53,
D18P186001,R1,D18P186001R1,54,,,BEPA,,D18P186001R1T54,
D18P186001,R1,D18P186001R1,55,1.614174062,,BEPA,,D18P186001R1T55,
D18P186001,R1,D18P186001R1,56,2.401576062,,,,D18P186001R1T56,
D18P186001,R1,D18P186001R1,57,2.992127562,,,,D18P186001R1T57,
D18P186001,R1,D18P186001R1,58,2.204725562,,,,D18P186001R1T58,
D18P186001,R1,D18P186001R1,59,4.96063275,,ABBA,,D18P186001R1T59,
D18P186001,R1,D18P186001R1,60,1.456693719,,ABBA,,D18P186001R1T60,
D18P186001,R1,D18P186001R1,61,,,BEPA,,D18P186001R1T61,
D18P186001,R1,D18P186001R1,62,0.905512281,,BEPA,,D18P186001R1T62,
D18P186001,R1,D18P186001R1,63,1.023622562,,BEPA,,D18P186001R1T63,
D18P186001,R1,D18P186001R1,64,1.732284438,,ABBA,,D18P186001R1T64,
D18P186001,R1,D18P186001R1,65,0.629921609,,BEPA,,D18P186001R1T65,
D18P186001,R1,D18P186001R1,66,6.65354675,,ABBA,,D18P186001R1T66,
D18P186001,R1,D18P186001R1,67,7.204728,,ABBA,34,D18P186001R1T67,
D18P186001,R1,D18P186001R1,68,1.259843219,,ABBA,,D18P186001R1T68,
D18P186001,R1,D18P186001R1,69,5.66929425,,ABBA,,D18P186001R1T69,
D18P186001,R1,D18P186001R1,70,7.6771695,,ABBA,28,D18P186001R1T70,
D18P186001,R1,D18P186001R1,71,0.5905515,,BEPA,,D18P186001R1T71,
D18P186001,R1,D18P186001R1,72,5.629924375,,ABBA,25,D18P186001R1T72,
D18P186001,R1,D18P186001R1,73,0.275590695,,BEPA,,D18P186001R1T73,
D18P186001,R1,D18P186001R1,74,0.314960805,,BEPA,,D18P186001R1T74,
D18P186001,R1,D18P186001R1,75,0.354330891,,BEPA,,D18P186001R1T75,
D18P186001,R1,D18P186001R1,76,0.472441219,,BEPA,,D18P186001R1T76,
D18P186001,R1,D18P186001R1,77,0.944882438,,ABBA,,D18P186001R1T77,
D18P186001,R1,D18P186001R1,78,0.393701,,BEPA,,D18P186001R1T78,
D18P186001,R1,D18P186001R1,79,5.433073875,,ABBA,,D18P186001R1T79,
D18P186001,R1,D18P186001R1,80,8.03150025,,ABBA,,D18P186001R1T80,
D18P186001,R1,D18P186001R1,81,,,ABBA,,D18P186001R1T81,
D18P186001,R1,D18P186001R1,82,0.157480402,,BEPA,,D18P186001R1T82,
D18P186001,R1,D18P186001R1,83,0.118110305,,BEPA,,D18P186001R1T83,
D18P186001,R1,D18P186001R1,84,0.118110305,,BEPA,,D18P186001R1T84,
D18P186001,R1,D18P186001R1,85,0.118110305,,BEPA,,D18P186001R1T85,
D18P186001,R1,D18P186001R1,86,0.118110305,,BEPA,,D18P186001R1T86,
D18P186001,R1,D18P186001R1,87,0.157480402,,BEPA,,D18P186001R1T87,
D18P186001,R1,D18P186001R1,88,0.275590695,,BEPA,,D18P186001R1T88,
D18P186001,R1,D18P186001R1,89,1.102362781,,BEPA,,D18P186001R1T89,
D18P186001,R1,D18P186001R1,90,0.314960805,,BEPA,,D18P186001R1T90,
D18P186001,R1,D18P186001R1,91,0.1968505,,BEPA,,D18P186001R1T91,
D18P186001,R1,D18P186001R1,92,0.629921609,,ABBA,,D18P186001R1T92,
D18P186001,R1,D18P186001R1,93,0.1968505,,ABBA,,D18P186001R1T93,
D18P186001,R1,D18P186001R1,94,0.393701,,ABBA,,D18P186001R1T94,
D18P186001,R1,D18P186001R1,95,0.314960805,,BEPA,,D18P186001R1T95,
D18P186001,R1,D18P186001R1,96,0.393701,,BEPA,,D18P186001R1T96,
D18P186001,R1,D18P186001R1,97,0.275590695,,BEPA,,D18P186001R1T97,
D18P186001,R1,D18P186001R1,98,0.393701,,BEPA,,D18P186001R1T98,
D18P186001,R1,D18P186001R1,99,0.354330891,,BEPA,,D18P186001R1T99,
D18P186001,R1,D18P186001R1,100,0.354330891,,BEPA,,D18P186001R1T100,
D18P186001,R2,D18P186001R2,1,6.3779565,,ABBA,,D18P186001R2T1,
D18P186001,R2,D18P186001R2,2,1.417323562,,BEPA,,D18P186001R2T2,
D18P186001,R2,D18P186001R2,3,,,BEPA,,D18P186001R2T3,
D18P186001,R2,D18P186001R2,4,,,BEPA,,D18P186001R2T4,
D18P186001,R2,D18P186001R2,5,0.314960805,,BEPA,,D18P186001R2T5,
D18P186001,R2,D18P186001R2,6,4.251970875,,ABBA,33,D18P186001R2T6,
D18P186001,R2,D18P186001R2,7,3.58267925,,ABBA,,D18P186001R2T7,
D18P186001,R2,D18P186001R2,8,,,,,D18P186001R2T8,
D18P186001,R2,D18P186001R2,9,2.5590565,,,,D18P186001R2T9,
D18P186001,R2,D18P186001R2,10,1.220473062,,,,D18P186001R2T10,
D18P186001,R2,D18P186001R2,11,1.614174062,,BEPA,,D18P186001R2T11,
D18P186001,R2,D18P186001R2,12,5.07874275,,ABBA,,D18P186001R2T12,
D18P186001,R2,D18P186001R2,13,4.1338605,,ABBA,,D18P186001R2T13,
D18P186001,R2,D18P186001R2,14,1.3779535,,ABBA,,D18P186001R2T14,
D18P186001,R2,D18P186001R2,15,5.433073875,,ABBA,35,D18P186001R2T15,
D18P186001,R2,D18P186001R2,16,4.37008125,,ABBA,35,D18P186001R2T16,
D18P186001,R2,D18P186001R2,17,1.220473062,,ABBA,,D18P186001R2T17,
D18P186001,R2,D18P186001R2,18,1.181103,,BEPA,,D18P186001R2T18,
D18P186001,R2,D18P186001R2,19,0.393701,,BEPA,,D18P186001R2T19,
D18P186001,R2,D18P186001R2,20,0.551181391,,BEPA,,D18P186001R2T20,
D18P186001,R2,D18P186001R2,21,2.834647125,,ABBA,29,D18P186001R2T21,
D18P186001,R2,D18P186001R2,22,,,BEPA,,D18P186001R2T22,
D18P186001,R2,D18P186001R2,23,1.259843219,,ABBA,,D18P186001R2T23,
D18P186001,R2,D18P186001R2,24,5.3149635,,ABBA,32,D18P186001R2T24,
D18P186001,R2,D18P186001R2,25,5.039372875,,ABBA,,D18P186001R2T25,
D18P186001,R2,D18P186001R2,26,,,BEPA,,D18P186001R2T26,
D18P186001,R2,D18P186001R2,27,0.9842525,,BEPA,,D18P186001R2T27,
D18P186001,R2,D18P186001R2,28,4.48819125,,ABBA,,D18P186001R2T28,
D18P186001,R2,D18P186001R2,29,0.393701,,BEPA,,D18P186001R2T29,
D18P186001,R2,D18P186001R2,30,0.551181391,,BEPA,,D18P186001R2T30,
D18P186001,R2,D18P186001R2,31,6.06299525,,ABBA,,D18P186001R2T31,
D18P186001,R2,D18P186001R2,32,5.55118425,,ABBA,,D18P186001R2T32,
D18P186001,R2,D18P186001R2,33,,,,,D18P186001R2T33,
D18P186001,R2,D18P186001R2,34,1.023622562,,,,D18P186001R2T34,
D18P186001,R2,D18P186001R2,35,0.905512281,,,,D18P186001R2T35,
D18P186001,R2,D18P186001R2,36,1.141732938,,,,D18P186001R2T36,
D18P186001,R2,D18P186001R2,37,4.76378225,,ABBA,37,D18P186001R2T37,
D18P186001,R2,D18P186001R2,38,5.94488525,,ABBA,,D18P186001R2T38,
D18P186001,R2,D18P186001R2,39,5.118113,,ABBA,,D18P186001R2T39,
D18P186001,R2,D18P186001R2,40,5.826774875,,ABBA,,D18P186001R2T40,
D18P186001,R2,D18P186001R2,41,5.039372875,,ABBA,,D18P186001R2T41,
D18P186001,R2,D18P186001R2,42,7.5590595,,ABBA,31,D18P186001R2T42,
D18P186001,R2,D18P186001R2,43,5.07874275,,ABBA,,D18P186001R2T43,
D18P186001,R2,D18P186001R2,44,4.842522375,,ABBA,,D18P186001R2T44,
D18P186001,R2,D18P186001R2,45,0.748031891,,ABBA,,D18P186001R2T45,
D18P186001,R2,D18P186001R2,46,1.417323562,,ABBA,,D18P186001R2T46,
D18P186001,R2,D18P186001R2,47,4.5275615,,ABBA,,D18P186001R2T47,
D18P186001,R2,D18P186001R2,48,1.102362781,,ABBA,,D18P186001R2T48,
D18P186001,R2,D18P186001R2,49,0.5905515,,ABBA,,D18P186001R2T49,
D18P186001,R2,D18P186001R2,50,5.27559325,,ABBA,,D18P186001R2T50,
D18P186001,R2,D18P186001R2,51,2.5590565,,BEPA,,D18P186001R2T51,
D18P186001,R2,D18P186001R2,52,,,,,D18P186001R2T52,
D18P186001,R2,D18P186001R2,53,,,BEPA,,D18P186001R2T53,
D18P186001,R2,D18P186001R2,54,,,BEPA,,D18P186001R2T54,
D18P186001,R2,D18P186001R2,55,1.653544125,,BEPA,,D18P186001R2T55,
D18P186001,R2,D18P186001R2,56,2.401576062,,,,D18P186001R2T56,
D18P186001,R2,D18P186001R2,57,3.149608,,,,D18P186001R2T57,
D18P186001,R2,D18P186001R2,58,2.440946125,,,,D18P186001R2T58,
D18P186001,R2,D18P186001R2,59,5.15748325,,ABBA,,D18P186001R2T59,
D18P186001,R2,D18P186001R2,60,1.653544125,,ABBA,,D18P186001R2T60,
D18P186001,R2,D18P186001R2,61,,,BEPA,,D18P186001R2T61,
D18P186001,R2,D18P186001R2,62,0.944882438,,BEPA,,D18P186001R2T62,
D18P186001,R2,D18P186001R2,63,1.102362781,,BEPA,,D18P186001R2T63,
D18P186001,R2,D18P186001R2,64,1.811024562,,ABBA,,D18P186001R2T64,
D18P186001,R2,D18P186001R2,65,0.708661781,,BEPA,,D18P186001R2T65,
D18P186001,R2,D18P186001R2,66,7.12598825,,ABBA,,D18P186001R2T66,
D18P186001,R2,D18P186001R2,67,7.51968925,,ABBA,36,D18P186001R2T67,
D18P186001,R2,D18P186001R2,68,1.3779535,,ABBA,,D18P186001R2T68,
D18P186001,R2,D18P186001R2,69,6.14173575,,ABBA,,D18P186001R2T69,
D18P186001,R2,D18P186001R2,70,8.89764275,,ABBA,30,D18P186001R2T70,
D18P186001,R2,D18P186001R2,71,0.787402,,BEPA,,D18P186001R2T71,
D18P186001,R2,D18P186001R2,72,5.826774875,,ABBA,27,D18P186001R2T72,
D18P186001,R2,D18P186001R2,73,0.354330891,,BEPA,,D18P186001R2T73,
D18P186001,R2,D18P186001R2,74,0.511811281,,BEPA,,D18P186001R2T74,
D18P186001,R2,D18P186001R2,75,0.433071109,,BEPA,,D18P186001R2T75,
D18P186001,R2,D18P186001R2,76,0.551181391,,BEPA,,D18P186001R2T76,
D18P186001,R2,D18P186001R2,77,1.062992719,,ABBA,,D18P186001R2T77,
D18P186001,R2,D18P186001R2,78,0.511811281,,BEPA,,D18P186001R2T78,
D18P186001,R2,D18P186001R2,79,5.826774875,,ABBA,,D18P186001R2T79,
D18P186001,R2,D18P186001R2,80,8.50394175,,ABBA,,D18P186001R2T80,
D18P186001,R2,D18P186001R2,81,,,ABBA,,D18P186001R2T81,
D18P186001,R2,D18P186001R2,82,0.157480402,,BEPA,,D18P186001R2T82,
D18P186001,R2,D18P186001R2,83,0.118110305,,BEPA,,D18P186001R2T83,
D18P186001,R2,D18P186001R2,84,,,BEPA,,D18P186001R2T84,
D18P186001,R2,D18P186001R2,85,,,BEPA,,D18P186001R2T85,
D18P186001,R2,D18P186001R2,86,0.118110305,,BEPA,,D18P186001R2T86,
D18P186001,R2,D18P186001R2,87,,,BEPA,,D18P186001R2T87,
D18P186001,R2,D18P186001R2,88,0.314960805,,BEPA,,D18P186001R2T88,
D18P186001,R2,D18P186001R2,89,,,BEPA,,D18P186001R2T89,
D18P186001,R2,D18P186001R2,90,0.433071109,,BEPA,,D18P186001R2T90,
D18P186001,R2,D18P186001R2,91,0.078740201,,BEPA,,D18P186001R2T91,
D18P186001,R2,D18P186001R2,92,0.708661781,,ABBA,,D18P186001R2T92,
D18P186001,R2,D18P186001R2,93,0.393701,,ABBA,,D18P186001R2T93,
D18P186001,R2,D18P186001R2,94,0.472441219,,ABBA,,D18P186001R2T94,
D18P186001,R2,D18P186001R2,95,,,BEPA,,D18P186001R2T95,
D18P186001,R2,D18P186001R2,96,0.393701,,BEPA,,D18P186001R2T96,
D18P186001,R2,D18P186001R2,97,0.393701,,BEPA,,D18P186001R2T97,
D18P186001,R2,D18P186001R2,98,0.393701,,BEPA,,D18P186001R2T98,
D18P186001,R2,D18P186001R2,99,0.472441219,,BEPA,,D18P186001R2T99,
D18P186001,R2,D18P186001R2,100,0.433071109,,BEPA,,D18P186001R2T100,
D18P186001,R3,D18P186001R3,1,6.92913775,,ABBA,,D18P186001R3T1,
D18P186001,R3,D18P186001R3,2,1.456693719,,BEPA,,D18P186001R3T2,
D18P186001,R3,D18P186001R3,3,,,BEPA,,D18P186001R3T3,
D18P186001,R3,D18P186001R3,4,,,BEPA,,D18P186001R3T4,
D18P186001,R3,D18P186001R3,5,,,BEPA,,D18P186001R3T5,
D18P186001,R3,D18P186001R3,6,4.48819125,,ABBA,37,D18P186001R3T6,
D18P186001,R3,D18P186001R3,7,3.858269875,,ABBA,,D18P186001R3T7,
D18P186001,R3,D18P186001R3,8,,,,,D18P186001R3T8,
D18P186001,R3,D18P186001R3,9,2.716536938,,,,D18P186001R3T9,
D18P186001,R3,D18P186001R3,10,,,,,D18P186001R3T10,
D18P186001,R3,D18P186001R3,11,1.535433938,,BEPA,,D18P186001R3T11,
D18P186001,R3,D18P186001R3,12,5.590554125,,ABBA,,D18P186001R3T12,
D18P186001,R3,D18P186001R3,13,4.37008125,,ABBA,,D18P186001R3T13,
D18P186001,R3,D18P186001R3,14,1.338583438,,ABBA,,D18P186001R3T14,
D18P186001,R3,D18P186001R3,15,5.984255125,,ABBA,39,D18P186001R3T15,
D18P186001,R3,D18P186001R3,16,4.88189225,,ABBA,39,D18P186001R3T16,
D18P186001,R3,D18P186001R3,17,1.338583438,,ABBA,,D18P186001R3T17,
D18P186001,R3,D18P186001R3,18,1.141732938,,BEPA,,D18P186001R3T18,
D18P186001,R3,D18P186001R3,19,,,BEPA,,D18P186001R3T19,
D18P186001,R3,D18P186001R3,20,0.669291719,,BEPA,,D18P186001R3T20,
D18P186001,R3,D18P186001R3,21,3.070867875,,ABBA,33,D18P186001R3T21,
D18P186001,R3,D18P186001R3,22,,,BEPA,,D18P186001R3T22,
D18P186001,R3,D18P186001R3,23,1.259843219,,ABBA,,D18P186001R3T23,
D18P186001,R3,D18P186001R3,24,5.905515,,ABBA,36,D18P186001R3T24,
D18P186001,R3,D18P186001R3,25,5.433073875,,ABBA,,D18P186001R3T25,
D18P186001,R3,D18P186001R3,26,,,BEPA,,D18P186001R3T26,
D18P186001,R3,D18P186001R3,27,1.023622562,,BEPA,,D18P186001R3T27,
D18P186001,R3,D18P186001R3,28,4.96063275,,ABBA,,D18P186001R3T28,
D18P186001,R3,D18P186001R3,29,,,BEPA,,D18P186001R3T29,
D18P186001,R3,D18P186001R3,30,0.511811281,,BEPA,,D18P186001R3T30,
D18P186001,R3,D18P186001R3,31,6.6141765,,ABBA,,D18P186001R3T31,
D18P186001,R3,D18P186001R3,32,6.06299525,,ABBA,,D18P186001R3T32,
D18P186001,R3,D18P186001R3,33,,,,,D18P186001R3T33,
D18P186001,R3,D18P186001R3,34,0.9842525,,,,D18P186001R3T34,
D18P186001,R3,D18P186001R3,35,,,,,D18P186001R3T35,
D18P186001,R3,D18P186001R3,36,1.220473062,,,,D18P186001R3T36,
D18P186001,R3,D18P186001R3,37,5.236223375,,ABBA,41,D18P186001R3T37,
D18P186001,R3,D18P186001R3,38,6.4960665,,ABBA,,D18P186001R3T38,
真正的答案
基本上，您编写的代码就像您用其他语言编写过很多但从未使用过的代码 R。由于所有嵌套循环，很难准确理解发生了什么，而且我没有 运行 你的代码（请参阅 here 了解如何可重复地包含数据）。
我认为虽然可以删除所有循环，例如用类似的东西替换第一个条件：
FVS_TreeInit$GD[
FVS_TreeInit$Plot_ID == "R8" &
FVS_TreeInit$StandPlotTree_ID == paste(FVS_TreeInit$Stand_ID,"R7T",FVS_TreeInit$Tree_ID,sep="")
] <- FVS_TreeInit$DBH
如果您使用 data.table（此处），它将是 significantly quicker，因为它的实现速度如此之快。
便宜又简单的答案
有一种既便宜又简单的方法可以在不重构所有代码的情况下加快您现有的速度，但它仍然会很慢。
问题是您一直在使用 df[i, "columnName"] 查找数据。这是一种非常缓慢的方法。看这个例子。
创建两个随机的大（10e6 行）字符串数据框：
# Random data function borrowed from: https://www.r-bloggers.com/2017/03/fast-data-lookups-in-r-dplyr-vs-data-table/
random_string_column <- function(n) {
stringi::stri_rand_strings(n = n, length = 8)
}
random_data_frame <- function(n) tibble(
col1 = random_string_column(n),
col2 = random_string_column(n)
)
df <- random_data_frame(10^6)
df2 <- random_data_frame(10^6)
head(df)
输出（df2会类似）：
# A tibble: 6 x 2
col1 col2
<chr> <chr>
1 NykzxkNO qGusGkMa
2 mgivTVky xlCtwEba
3 zar6FbC4 dOYeQPfg
4 89fKLJFv RPsP2CZc
5 V2FH3Smt ObToBJFC
6 gXa0j5XZ QYKK6hD2
现在让我们按照您的方式进行 1000 次随机查找：
rand_indeces <- sample(c(1:nrow(df), 1000))
system.time(
for(i in rand_indeces) {
if(df[i, "col2"]=="aaaaa") { #it never will
break
}
}
)
输出：
user system elapsed
138.58 0.23 140.72
两分钟多了。现在让我们看看另一个数据框（新的 df 所以我们可以看到它不是因为缓存），用 df2$col2[i] == "aaaaa".
替换 df[i, "col2"]=="aaaaa" 语法
system.time(
for(i in rand_indeces) {
if(df2$col2[i] == "aaaaa") { #it never will
break
}
}
)
输出：
user system elapsed
4.70 0.00 4.75
这应该可以为您节省不少时间。但是，从根本上说，在 R 中以这种方式进行查找很慢（参见 here and ）。
如果您必须进行单独的查找，您可以考虑创建一个散列 table，或者通过第一个 link 中的 environments 或者通过一些较新的方法 here.我认为 collections 包看起来很有趣 - 还没有使用它。
但是，到目前为止，最好的办法是矢量化您的代码，如本回复顶部所述。
（扰流板：跳转到 Double Merge 部分以获得比嵌套循环快 150 倍的结果。我希望这种速度改进随着数据的增加而变得更好。）
中级改进
首先，您的内部循环都可以从 for 循环中减少，从而变得更快，也许会改变
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
到
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
f <- FVS_TreeInit[,"StandPlotTree_ID"] == find
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
值得注意的是，我们已经从迭代 for 循环更改为 向量化 比较，这在 R 中往往要快得多。但是，应该注意我们对几个 "R*" 做同样的事情，所以我们可以大大减少你的外层循环；其中大部分是代码高尔夫（减少字符数来做同样的事情），但总的来说不要重复你自己：如果你有相同的（相同的）代码多个地方，你增加了改变一个而不是所有其他地方的机会。
在那个主题中，到目前为止，我可以通过使用这个得到相同的结果：
# loop through FVS_TreeInit and add previous diameter increment identifier to current row
for (i in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[i,"Plot_ID"] == "R8"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R7T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R7"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R6T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R6"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R5T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R5"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R4T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R4"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R3T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R3"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R2T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R2"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R1T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R1"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else {
find <- NA
}
FVS_TreeInit[i,"GD"] <- NA
if (!is.na(find)) {
f <- FVS_TreeInit[,"StandPlotTree_ID"] == find
if (any(f)) FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
}
if (i%%1000 == 0){
print(paste(i," rows iterated...",sep=""))
}
}
好的，那更短了，缩小到真正重要的部分。值得注意的是（对我而言），我们在这里根据某些特定值有效地将数据“连接”在一起。我认为我们可以通过更简单的步骤（合并）完成所有这些工作。
双重合并
这实际上只是根据三个 ID：Stand_ID、Tree_ID 和 Plot_ID 将数据自身连接回去，其中后者是一个时间组件（因此需要为自加入进行调整）。
基于您的 R* 值反映时间的想法，我认为我们可以安全地假设 R(n) 应该分配给 R(n+1)，所以让我们以编程方式创建这个中间框架：
prevPlot <- data.frame(prev_Plot_ID = unique(FVS_TreeInit$Plot_ID))
prevPlot$Plot_ID <- paste0("R", as.integer(gsub("R", "", prevPlot$prev_Plot_ID))-1)
prevPlot <- subset(prevPlot, prev_Plot_ID != "R0")
prevPlot
# prev_Plot_ID Plot_ID
# 2 R1 R0
# 3 R2 R1
# 4 R3 R2
我们将 merge/join 它放到原来的 FVS_Tree_Init 上以识别“先前的 Plot_ID”，然后合并回自身以引入新值。
prevValues <- subset(merge(FVS_TreeInit, prevPlot, by = "Plot_ID", all.x = TRUE),
select = c(Stand_ID, prev_Plot_ID, Tree_ID, DBH))
names(prevValues)[4] <- "GD"
head(prevValues)
# Stand_ID prev_Plot_ID Tree_ID GD
# 1 D18P186001 R1 39 3.9370100
# 2 D18P186001 R1 40 4.2519709
# 3 D18P186001 R1 1 4.7637823
# 4 D18P186001 R1 2 1.2598432
# 5 D18P186001 R1 3 0.2755907
# 6 D18P186001 R1 4 0.6299216
new_FVS_TreeInit <- merge(
FVS_TreeInit, prevValues,
by.x = c("Stand_ID", "Plot_ID", "Tree_ID"),
by.y = c("Stand_ID", "prev_Plot_ID", "Tree_ID"),
all.x = TRUE)
identical(expected, new_FVS_TreeInit[names(expected)])
# [1] TRUE
其中 expected 是嵌套 for 循环的输出，new_FVS_TreeInit[names(expected)] 只是确保所有列顺序相同的一个步骤（与 identical，数据处理不需要）。
仅供参考，这突出了您使用连接字符串作为单数键的矛盾（一种相当 Excel- 进行连接的方法......它确实有效）前提是 merging/joining 一个或多个键上的数据。一旦您放弃单键 Excel 方法并认识到 一个或多个 也同样有效——并且更容易阅读和维护——我认为转变观念合并可以改变编程。有关联接“演算”的更多讨论，请参阅 How to join (merge) data frames (inner, outer, left, right) and What's the difference between INNER JOIN, LEFT JOIN, RIGHT JOIN and FULL JOIN?。
基准
使用此示例数据，执行速度大约 150 倍。
bench::mark(
forloop = {
# your code
},
doublemerge = {
# my code
}
)
# Warning: Some expressions had a GC in every iteration; so filtering is disabled.
# # A tibble: 2 x 13
# expression min median `itr/sec` mem_alloc `gc/sec` n_itr n_gc total_time result memory time gc
# <bch:expr> <bch:tm> <bch:tm> <dbl> <bch:byt> <dbl> <int> <dbl> <bch:tm> <list> <list> <list> <list>
# 1 forloop 272.4ms 284.86ms 3.51 656KB 7.02 2 4 570ms <df [318 x 11]> <Rprofmem ~ <bench_t~ <tibble ~
# 2 doublemerge 1.4ms 1.55ms 528. 346KB 1.99 265 1 502ms <df [318 x 11]> <Rprofmem ~ <bench_t~ <tibble ~
单程data.table解决方案：
这可以在依次按 Stand_ID、Tree_ID、Plot_ID 排序后一次性完成（在我的机器上约 6 毫秒，包括读取数据）。
首先，我将示例数据字符串复制到对象 data。
library(data.table)
dt <- fread(input = data) # read into a data.table
setorder(dt[, `:=`(Plot_ID = as.integer(substr(Plot_ID, 2, 2)), DG = as.numeric(DG))], # convert Plot_ID to an integer and DG to numeric (from logical)
# order by Stand_ID, Tree_ID, Plot_ID
Stand_ID, Tree_ID, Plot_ID)[
# create a temporary column indicating the rows that will pull DBH from the previous row
, pull := c(FALSE, diff(Plot_ID) == 1L) & Stand_ID == shift(Stand_ID) & Tree_ID == shift(Tree_ID)][
# pull DBH from the previous row where appropriate
pull == TRUE, DG := shift(dt$DBH)[dt$pull]][,
# clean up
`:=`(pull = NULL, Plot_ID = paste0("R", Plot_ID))]
我相信这可以用更少的代码以更简单的方式重构。看起来您正在对数据进行连接，其中查找键是当前 Plot_ID 的修改版本，其中“R#”应链接到“R#-1T”。
在 R 中执行此操作的更惯用的方法是为所有行创建一次适当的键，然后根据该查找键进行连接。这将比您的循环方法快几个数量级，因为它只需要编译查找代码一次而不是 650,000 次，并且连接针对速度进行了优化。请参阅此页面以深入解释为什么向量化在 R 中有如此大的帮助：https://www.noamross.net/archives/2014-04-16-vectorization-in-r-why/
我在下面使用 dplyr 执行此操作，但该方法在 base r 中的工作方式非常相似。
library(dplyr)
FVS_TreeInit %>%
# create key for all rows
mutate(prior_key = paste0(Stand_ID, "R",
parse_number(Plot_ID)-1, "T", # will be nonsense R-1T when starts from R0, that's ok,
Tree_ID)) %>% # it just won't replace GD there
# Use that key to join to a version of the data with just StandPlotTree_ID and
# DBH, which I rename as GD. Match `prior_key` with `StandPlotTree_ID` there.
left_join(FVS_TreeInit %>% transmute(StandPlotTree_ID, GD = DBH),
by = c("prior_key" = "StandPlotTree_ID"))
基准
您的原始代码在提供的示例数据上 运行 花费了 3.7 秒。我建议的更换用了 0.027 秒，大约是更换速度的 140 倍。我预计随着数据的增加，时差会变得更大。
我准备了一个例程，将一个单元格（“DBH”）的值复制到另一行中的另一个单元格（“DG”），其中被复制到的行紧接在被复制的行之后（及时，不一定在table）。
PlotID 'R8' 表示第 8 个测量值，'R7' 第 7 个，依此类推。
该循环旨在检查该行是否具有先前的测量值，如果有，则记录该测量值的唯一行标识符。然后它使用前面的行标识符来搜索要通过第二个嵌套循环复制的前面的记录。
问题是我使用的完整数据集包含超过 650,000 条记录。我知道这种结构会有些低效，但是当我 运行 它时 运行 时间每小时仅移动大约 10,000 条记录。有什么提高计算效率的建议吗？
# sort TreeInit Table to facilitate desired iteration
FVS_TreeInit$StandPlotTree_ID <- paste(FVS_TreeInit$StandPlot_ID,"T",FVS_TreeInit$Tree_ID,sep="")
# loop through FVS_TreeInit and add previous diameter increment identifier to current row
for (i in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[i,"Plot_ID"] == "R8"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R7T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R7"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R6T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R6"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R5T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R5"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R4T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R4"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R3T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R3"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R2T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R2"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R1T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else if (FVS_TreeInit[i,"Plot_ID"] == "R1"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
} else{
FVS_TreeInit[i,"GD"] <- NA
}
if (i%%1000 == 0){
print(paste(i," rows iterated...",sep=""))
}
}
}
一些样本数据（.csv）：
Stand_ID,Plot_ID,StandPlot_ID,Tree_ID,DBH,DG,Species,Age,StandPlotTree_ID,StandPlotTree_ID_Past
D18P186001,R0,D18P186001R0,1,4.76378225,,ABBA,,D18P186001R0T1,
D18P186001,R0,D18P186001R0,2,1.259843219,,BEPA,,D18P186001R0T2,
D18P186001,R0,D18P186001R0,3,0.275590695,,BEPA,,D18P186001R0T3,
D18P186001,R0,D18P186001R0,4,0.629921609,,BEPA,,D18P186001R0T4,
D18P186001,R0,D18P186001R0,5,0.1968505,,BEPA,,D18P186001R0T5,
D18P186001,R0,D18P186001R0,6,3.70078925,,ABBA,26,D18P186001R0T6,
D18P186001,R0,D18P186001R0,7,2.834647125,,ABBA,,D18P186001R0T7,
D18P186001,R0,D18P186001R0,8,1.259843219,,,,D18P186001R0T8,
D18P186001,R0,D18P186001R0,9,2.283465875,,,,D18P186001R0T9,
D18P186001,R0,D18P186001R0,10,1.181103,,,,D18P186001R0T10,
D18P186001,R0,D18P186001R0,11,1.535433938,,BEPA,,D18P186001R0T11,
D18P186001,R0,D18P186001R0,12,4.212600625,,ABBA,,D18P186001R0T12,
D18P186001,R0,D18P186001R0,13,3.30708825,,ABBA,,D18P186001R0T13,
D18P186001,R0,D18P186001R0,14,1.259843219,,ABBA,,D18P186001R0T14,
D18P186001,R0,D18P186001R0,15,4.212600625,,ABBA,28,D18P186001R0T15,
D18P186001,R0,D18P186001R0,16,3.149608,,ABBA,28,D18P186001R0T16,
D18P186001,R0,D18P186001R0,17,1.181103,,ABBA,,D18P186001R0T17,
D18P186001,R0,D18P186001R0,18,0.944882438,,BEPA,,D18P186001R0T18,
D18P186001,R0,D18P186001R0,19,0.393701,,BEPA,,D18P186001R0T19,
D18P186001,R0,D18P186001R0,20,0.314960805,,BEPA,,D18P186001R0T20,
D18P186001,R0,D18P186001R0,21,2.362206,,ABBA,22,D18P186001R0T21,
D18P186001,R0,D18P186001R0,22,0.314960805,,BEPA,,D18P186001R0T22,
D18P186001,R0,D18P186001R0,23,1.062992719,,ABBA,,D18P186001R0T23,
D18P186001,R0,D18P186001R0,24,3.93701,,ABBA,25,D18P186001R0T24,
D18P186001,R0,D18P186001R0,25,4.76378225,,ABBA,,D18P186001R0T25,
D18P186001,R0,D18P186001R0,26,0.748031891,,BEPA,,D18P186001R0T26,
D18P186001,R0,D18P186001R0,27,0.944882438,,BEPA,,D18P186001R0T27,
D18P186001,R0,D18P186001R0,28,3.267718375,,ABBA,,D18P186001R0T28,
D18P186001,R0,D18P186001R0,29,0.314960805,,BEPA,,D18P186001R0T29,
D18P186001,R0,D18P186001R0,30,0.472441219,,BEPA,,D18P186001R0T30,
D18P186001,R0,D18P186001R0,31,4.606301625,,ABBA,,D18P186001R0T31,
D18P186001,R0,D18P186001R0,32,4.055120375,,ABBA,,D18P186001R0T32,
D18P186001,R0,D18P186001R0,33,0.433071109,,,,D18P186001R0T33,
D18P186001,R0,D18P186001R0,34,1.023622562,,,,D18P186001R0T34,
D18P186001,R0,D18P186001R0,35,0.5905515,,,,D18P186001R0T35,
D18P186001,R0,D18P186001R0,36,0.748031891,,,,D18P186001R0T36,
D18P186001,R0,D18P186001R0,37,3.425198625,,ABBA,30,D18P186001R0T37,
D18P186001,R0,D18P186001R0,38,4.29134075,,ABBA,,D18P186001R0T38,
D18P186001,R0,D18P186001R0,39,3.93701,,ABBA,,D18P186001R0T39,
D18P186001,R0,D18P186001R0,40,4.251970875,,ABBA,,D18P186001R0T40,
D18P186001,R0,D18P186001R0,41,4.09449025,,ABBA,,D18P186001R0T41,
D18P186001,R0,D18P186001R0,42,6.181105625,,ABBA,24,D18P186001R0T42,
D18P186001,R0,D18P186001R0,43,3.858269875,,ABBA,,D18P186001R0T43,
D18P186001,R0,D18P186001R0,44,3.7401595,,ABBA,,D18P186001R0T44,
D18P186001,R0,D18P186001R0,45,0.5905515,,ABBA,,D18P186001R0T45,
D18P186001,R0,D18P186001R0,46,1.259843219,,ABBA,,D18P186001R0T46,
D18P186001,R0,D18P186001R0,47,3.89763975,,ABBA,,D18P186001R0T47,
D18P186001,R0,D18P186001R0,48,0.708661781,,ABBA,,D18P186001R0T48,
D18P186001,R0,D18P186001R0,49,0.314960805,,ABBA,,D18P186001R0T49,
D18P186001,R0,D18P186001R0,50,4.1338605,,ABBA,,D18P186001R0T50,
D18P186001,R0,D18P186001R0,51,1.889764875,,BEPA,,D18P186001R0T51,
D18P186001,R0,D18P186001R0,52,0.314960805,,,,D18P186001R0T52,
D18P186001,R0,D18P186001R0,53,0.629921609,,BEPA,,D18P186001R0T53,
D18P186001,R0,D18P186001R0,54,0.354330891,,BEPA,,D18P186001R0T54,
D18P186001,R0,D18P186001R0,55,1.456693719,,BEPA,,D18P186001R0T55,
D18P186001,R0,D18P186001R0,56,2.047245125,,,,D18P186001R0T56,
D18P186001,R0,D18P186001R0,57,2.480316375,,,,D18P186001R0T57,
D18P186001,R0,D18P186001R0,58,1.968505,,,,D18P186001R0T58,
D18P186001,R0,D18P186001R0,59,3.93701,,ABBA,,D18P186001R0T59,
D18P186001,R0,D18P186001R0,60,1.102362781,,ABBA,,D18P186001R0T60,
D18P186001,R0,D18P186001R0,61,0.472441219,,BEPA,,D18P186001R0T61,
D18P186001,R0,D18P186001R0,62,0.905512281,,BEPA,,D18P186001R0T62,
D18P186001,R0,D18P186001R0,63,0.944882438,,BEPA,,D18P186001R0T63,
D18P186001,R0,D18P186001R0,64,1.574804,,ABBA,,D18P186001R0T64,
D18P186001,R0,D18P186001R0,65,0.472441219,,BEPA,,D18P186001R0T65,
D18P186001,R0,D18P186001R0,66,5.15748325,,ABBA,,D18P186001R0T66,
D18P186001,R0,D18P186001R0,67,5.35433375,,ABBA,29,D18P186001R0T67,
D18P186001,R0,D18P186001R0,68,0.511811281,,ABBA,,D18P186001R0T68,
D18P186001,R0,D18P186001R0,69,4.68504175,,ABBA,,D18P186001R0T69,
D18P186001,R0,D18P186001R0,70,6.8897675,,ABBA,23,D18P186001R0T70,
D18P186001,R0,D18P186001R0,71,0.275590695,,BEPA,,D18P186001R0T71,
D18P186001,R0,D18P186001R0,72,4.606301625,,ABBA,20,D18P186001R0T72,
D18P186001,R0,D18P186001R0,73,0.511811281,,BEPA,,D18P186001R0T73,
D18P186001,R0,D18P186001R0,74,0.275590695,,BEPA,,D18P186001R0T74,
D18P186001,R0,D18P186001R0,75,0.393701,,BEPA,,D18P186001R0T75,
D18P186001,R0,D18P186001R0,76,0.472441219,,BEPA,,D18P186001R0T76,
D18P186001,R0,D18P186001R0,77,0.669291719,,ABBA,,D18P186001R0T77,
D18P186001,R0,D18P186001R0,78,0.354330891,,BEPA,,D18P186001R0T78,
D18P186001,R0,D18P186001R0,79,4.409451125,,ABBA,,D18P186001R0T79,
D18P186001,R0,D18P186001R0,80,7.12598825,,ABBA,,D18P186001R0T80,
D18P186001,R0,D18P186001R0,81,4.724412,,ABBA,,D18P186001R0T81,
D18P186001,R1,D18P186001R1,1,5.86614475,,ABBA,,D18P186001R1T1,
D18P186001,R1,D18P186001R1,2,1.3779535,,BEPA,,D18P186001R1T2,
D18P186001,R1,D18P186001R1,3,,,BEPA,,D18P186001R1T3,
D18P186001,R1,D18P186001R1,4,,,BEPA,,D18P186001R1T4,
D18P186001,R1,D18P186001R1,5,0.236220609,,BEPA,,D18P186001R1T5,
D18P186001,R1,D18P186001R1,6,4.1338605,,ABBA,31,D18P186001R1T6,
D18P186001,R1,D18P186001R1,7,3.425198625,,ABBA,,D18P186001R1T7,
D18P186001,R1,D18P186001R1,9,2.480316375,,,,D18P186001R1T9,
D18P186001,R1,D18P186001R1,10,1.259843219,,,,D18P186001R1T10,
D18P186001,R1,D18P186001R1,11,1.574804,,BEPA,,D18P186001R1T11,
D18P186001,R1,D18P186001R1,12,4.88189225,,ABBA,,D18P186001R1T12,
D18P186001,R1,D18P186001R1,13,3.89763975,,ABBA,,D18P186001R1T13,
D18P186001,R1,D18P186001R1,14,1.3779535,,ABBA,,D18P186001R1T14,
D18P186001,R1,D18P186001R1,15,5.196853125,,ABBA,33,D18P186001R1T15,
D18P186001,R1,D18P186001R1,16,4.055120375,,ABBA,33,D18P186001R1T16,
D18P186001,R1,D18P186001R1,17,1.220473062,,ABBA,,D18P186001R1T17,
D18P186001,R1,D18P186001R1,18,0.9842525,,BEPA,,D18P186001R1T18,
D18P186001,R1,D18P186001R1,19,0.393701,,BEPA,,D18P186001R1T19,
D18P186001,R1,D18P186001R1,20,0.472441219,,BEPA,,D18P186001R1T20,
D18P186001,R1,D18P186001R1,21,2.716536938,,ABBA,27,D18P186001R1T21,
D18P186001,R1,D18P186001R1,22,,,BEPA,,D18P186001R1T22,
D18P186001,R1,D18P186001R1,23,1.181103,,ABBA,,D18P186001R1T23,
D18P186001,R1,D18P186001R1,24,5.000002625,,ABBA,30,D18P186001R1T24,
D18P186001,R1,D18P186001R1,25,4.88189225,,ABBA,,D18P186001R1T25,
D18P186001,R1,D18P186001R1,26,,,BEPA,,D18P186001R1T26,
D18P186001,R1,D18P186001R1,27,1.023622562,,BEPA,,D18P186001R1T27,
D18P186001,R1,D18P186001R1,28,4.330711,,ABBA,,D18P186001R1T28,
D18P186001,R1,D18P186001R1,29,0.354330891,,BEPA,,D18P186001R1T29,
D18P186001,R1,D18P186001R1,30,0.511811281,,BEPA,,D18P186001R1T30,
D18P186001,R1,D18P186001R1,31,5.74803475,,ABBA,,D18P186001R1T31,
D18P186001,R1,D18P186001R1,32,5.27559325,,ABBA,,D18P186001R1T32,
D18P186001,R1,D18P186001R1,33,,,,,D18P186001R1T33,
D18P186001,R1,D18P186001R1,34,0.944882438,,,,D18P186001R1T34,
D18P186001,R1,D18P186001R1,35,0.669291719,,,,D18P186001R1T35,
D18P186001,R1,D18P186001R1,36,1.062992719,,,,D18P186001R1T36,
D18P186001,R1,D18P186001R1,37,4.5275615,,ABBA,35,D18P186001R1T37,
D18P186001,R1,D18P186001R1,38,5.629924375,,ABBA,,D18P186001R1T38,
D18P186001,R1,D18P186001R1,39,4.88189225,,ABBA,,D18P186001R1T39,
D18P186001,R1,D18P186001R1,40,5.47244375,,ABBA,,D18P186001R1T40,
D18P186001,R1,D18P186001R1,41,4.842522375,,ABBA,,D18P186001R1T41,
D18P186001,R1,D18P186001R1,42,7.204728,,ABBA,29,D18P186001R1T42,
D18P186001,R1,D18P186001R1,43,4.842522375,,ABBA,,D18P186001R1T43,
D18P186001,R1,D18P186001R1,44,4.606301625,,ABBA,,D18P186001R1T44,
D18P186001,R1,D18P186001R1,45,0.748031891,,ABBA,,D18P186001R1T45,
D18P186001,R1,D18P186001R1,46,1.3779535,,ABBA,,D18P186001R1T46,
D18P186001,R1,D18P186001R1,47,4.37008125,,ABBA,,D18P186001R1T47,
D18P186001,R1,D18P186001R1,48,0.9842525,,ABBA,,D18P186001R1T48,
D18P186001,R1,D18P186001R1,49,0.354330891,,ABBA,,D18P186001R1T49,
D18P186001,R1,D18P186001R1,50,5.07874275,,ABBA,,D18P186001R1T50,
D18P186001,R1,D18P186001R1,51,2.401576062,,BEPA,,D18P186001R1T51,
D18P186001,R1,D18P186001R1,52,,,,,D18P186001R1T52,
D18P186001,R1,D18P186001R1,53,,,BEPA,,D18P186001R1T53,
D18P186001,R1,D18P186001R1,54,,,BEPA,,D18P186001R1T54,
D18P186001,R1,D18P186001R1,55,1.614174062,,BEPA,,D18P186001R1T55,
D18P186001,R1,D18P186001R1,56,2.401576062,,,,D18P186001R1T56,
D18P186001,R1,D18P186001R1,57,2.992127562,,,,D18P186001R1T57,
D18P186001,R1,D18P186001R1,58,2.204725562,,,,D18P186001R1T58,
D18P186001,R1,D18P186001R1,59,4.96063275,,ABBA,,D18P186001R1T59,
D18P186001,R1,D18P186001R1,60,1.456693719,,ABBA,,D18P186001R1T60,
D18P186001,R1,D18P186001R1,61,,,BEPA,,D18P186001R1T61,
D18P186001,R1,D18P186001R1,62,0.905512281,,BEPA,,D18P186001R1T62,
D18P186001,R1,D18P186001R1,63,1.023622562,,BEPA,,D18P186001R1T63,
D18P186001,R1,D18P186001R1,64,1.732284438,,ABBA,,D18P186001R1T64,
D18P186001,R1,D18P186001R1,65,0.629921609,,BEPA,,D18P186001R1T65,
D18P186001,R1,D18P186001R1,66,6.65354675,,ABBA,,D18P186001R1T66,
D18P186001,R1,D18P186001R1,67,7.204728,,ABBA,34,D18P186001R1T67,
D18P186001,R1,D18P186001R1,68,1.259843219,,ABBA,,D18P186001R1T68,
D18P186001,R1,D18P186001R1,69,5.66929425,,ABBA,,D18P186001R1T69,
D18P186001,R1,D18P186001R1,70,7.6771695,,ABBA,28,D18P186001R1T70,
D18P186001,R1,D18P186001R1,71,0.5905515,,BEPA,,D18P186001R1T71,
D18P186001,R1,D18P186001R1,72,5.629924375,,ABBA,25,D18P186001R1T72,
D18P186001,R1,D18P186001R1,73,0.275590695,,BEPA,,D18P186001R1T73,
D18P186001,R1,D18P186001R1,74,0.314960805,,BEPA,,D18P186001R1T74,
D18P186001,R1,D18P186001R1,75,0.354330891,,BEPA,,D18P186001R1T75,
D18P186001,R1,D18P186001R1,76,0.472441219,,BEPA,,D18P186001R1T76,
D18P186001,R1,D18P186001R1,77,0.944882438,,ABBA,,D18P186001R1T77,
D18P186001,R1,D18P186001R1,78,0.393701,,BEPA,,D18P186001R1T78,
D18P186001,R1,D18P186001R1,79,5.433073875,,ABBA,,D18P186001R1T79,
D18P186001,R1,D18P186001R1,80,8.03150025,,ABBA,,D18P186001R1T80,
D18P186001,R1,D18P186001R1,81,,,ABBA,,D18P186001R1T81,
D18P186001,R1,D18P186001R1,82,0.157480402,,BEPA,,D18P186001R1T82,
D18P186001,R1,D18P186001R1,83,0.118110305,,BEPA,,D18P186001R1T83,
D18P186001,R1,D18P186001R1,84,0.118110305,,BEPA,,D18P186001R1T84,
D18P186001,R1,D18P186001R1,85,0.118110305,,BEPA,,D18P186001R1T85,
D18P186001,R1,D18P186001R1,86,0.118110305,,BEPA,,D18P186001R1T86,
D18P186001,R1,D18P186001R1,87,0.157480402,,BEPA,,D18P186001R1T87,
D18P186001,R1,D18P186001R1,88,0.275590695,,BEPA,,D18P186001R1T88,
D18P186001,R1,D18P186001R1,89,1.102362781,,BEPA,,D18P186001R1T89,
D18P186001,R1,D18P186001R1,90,0.314960805,,BEPA,,D18P186001R1T90,
D18P186001,R1,D18P186001R1,91,0.1968505,,BEPA,,D18P186001R1T91,
D18P186001,R1,D18P186001R1,92,0.629921609,,ABBA,,D18P186001R1T92,
D18P186001,R1,D18P186001R1,93,0.1968505,,ABBA,,D18P186001R1T93,
D18P186001,R1,D18P186001R1,94,0.393701,,ABBA,,D18P186001R1T94,
D18P186001,R1,D18P186001R1,95,0.314960805,,BEPA,,D18P186001R1T95,
D18P186001,R1,D18P186001R1,96,0.393701,,BEPA,,D18P186001R1T96,
D18P186001,R1,D18P186001R1,97,0.275590695,,BEPA,,D18P186001R1T97,
D18P186001,R1,D18P186001R1,98,0.393701,,BEPA,,D18P186001R1T98,
D18P186001,R1,D18P186001R1,99,0.354330891,,BEPA,,D18P186001R1T99,
D18P186001,R1,D18P186001R1,100,0.354330891,,BEPA,,D18P186001R1T100,
D18P186001,R2,D18P186001R2,1,6.3779565,,ABBA,,D18P186001R2T1,
D18P186001,R2,D18P186001R2,2,1.417323562,,BEPA,,D18P186001R2T2,
D18P186001,R2,D18P186001R2,3,,,BEPA,,D18P186001R2T3,
D18P186001,R2,D18P186001R2,4,,,BEPA,,D18P186001R2T4,
D18P186001,R2,D18P186001R2,5,0.314960805,,BEPA,,D18P186001R2T5,
D18P186001,R2,D18P186001R2,6,4.251970875,,ABBA,33,D18P186001R2T6,
D18P186001,R2,D18P186001R2,7,3.58267925,,ABBA,,D18P186001R2T7,
D18P186001,R2,D18P186001R2,8,,,,,D18P186001R2T8,
D18P186001,R2,D18P186001R2,9,2.5590565,,,,D18P186001R2T9,
D18P186001,R2,D18P186001R2,10,1.220473062,,,,D18P186001R2T10,
D18P186001,R2,D18P186001R2,11,1.614174062,,BEPA,,D18P186001R2T11,
D18P186001,R2,D18P186001R2,12,5.07874275,,ABBA,,D18P186001R2T12,
D18P186001,R2,D18P186001R2,13,4.1338605,,ABBA,,D18P186001R2T13,
D18P186001,R2,D18P186001R2,14,1.3779535,,ABBA,,D18P186001R2T14,
D18P186001,R2,D18P186001R2,15,5.433073875,,ABBA,35,D18P186001R2T15,
D18P186001,R2,D18P186001R2,16,4.37008125,,ABBA,35,D18P186001R2T16,
D18P186001,R2,D18P186001R2,17,1.220473062,,ABBA,,D18P186001R2T17,
D18P186001,R2,D18P186001R2,18,1.181103,,BEPA,,D18P186001R2T18,
D18P186001,R2,D18P186001R2,19,0.393701,,BEPA,,D18P186001R2T19,
D18P186001,R2,D18P186001R2,20,0.551181391,,BEPA,,D18P186001R2T20,
D18P186001,R2,D18P186001R2,21,2.834647125,,ABBA,29,D18P186001R2T21,
D18P186001,R2,D18P186001R2,22,,,BEPA,,D18P186001R2T22,
D18P186001,R2,D18P186001R2,23,1.259843219,,ABBA,,D18P186001R2T23,
D18P186001,R2,D18P186001R2,24,5.3149635,,ABBA,32,D18P186001R2T24,
D18P186001,R2,D18P186001R2,25,5.039372875,,ABBA,,D18P186001R2T25,
D18P186001,R2,D18P186001R2,26,,,BEPA,,D18P186001R2T26,
D18P186001,R2,D18P186001R2,27,0.9842525,,BEPA,,D18P186001R2T27,
D18P186001,R2,D18P186001R2,28,4.48819125,,ABBA,,D18P186001R2T28,
D18P186001,R2,D18P186001R2,29,0.393701,,BEPA,,D18P186001R2T29,
D18P186001,R2,D18P186001R2,30,0.551181391,,BEPA,,D18P186001R2T30,
D18P186001,R2,D18P186001R2,31,6.06299525,,ABBA,,D18P186001R2T31,
D18P186001,R2,D18P186001R2,32,5.55118425,,ABBA,,D18P186001R2T32,
D18P186001,R2,D18P186001R2,33,,,,,D18P186001R2T33,
D18P186001,R2,D18P186001R2,34,1.023622562,,,,D18P186001R2T34,
D18P186001,R2,D18P186001R2,35,0.905512281,,,,D18P186001R2T35,
D18P186001,R2,D18P186001R2,36,1.141732938,,,,D18P186001R2T36,
D18P186001,R2,D18P186001R2,37,4.76378225,,ABBA,37,D18P186001R2T37,
D18P186001,R2,D18P186001R2,38,5.94488525,,ABBA,,D18P186001R2T38,
D18P186001,R2,D18P186001R2,39,5.118113,,ABBA,,D18P186001R2T39,
D18P186001,R2,D18P186001R2,40,5.826774875,,ABBA,,D18P186001R2T40,
D18P186001,R2,D18P186001R2,41,5.039372875,,ABBA,,D18P186001R2T41,
D18P186001,R2,D18P186001R2,42,7.5590595,,ABBA,31,D18P186001R2T42,
D18P186001,R2,D18P186001R2,43,5.07874275,,ABBA,,D18P186001R2T43,
D18P186001,R2,D18P186001R2,44,4.842522375,,ABBA,,D18P186001R2T44,
D18P186001,R2,D18P186001R2,45,0.748031891,,ABBA,,D18P186001R2T45,
D18P186001,R2,D18P186001R2,46,1.417323562,,ABBA,,D18P186001R2T46,
D18P186001,R2,D18P186001R2,47,4.5275615,,ABBA,,D18P186001R2T47,
D18P186001,R2,D18P186001R2,48,1.102362781,,ABBA,,D18P186001R2T48,
D18P186001,R2,D18P186001R2,49,0.5905515,,ABBA,,D18P186001R2T49,
D18P186001,R2,D18P186001R2,50,5.27559325,,ABBA,,D18P186001R2T50,
D18P186001,R2,D18P186001R2,51,2.5590565,,BEPA,,D18P186001R2T51,
D18P186001,R2,D18P186001R2,52,,,,,D18P186001R2T52,
D18P186001,R2,D18P186001R2,53,,,BEPA,,D18P186001R2T53,
D18P186001,R2,D18P186001R2,54,,,BEPA,,D18P186001R2T54,
D18P186001,R2,D18P186001R2,55,1.653544125,,BEPA,,D18P186001R2T55,
D18P186001,R2,D18P186001R2,56,2.401576062,,,,D18P186001R2T56,
D18P186001,R2,D18P186001R2,57,3.149608,,,,D18P186001R2T57,
D18P186001,R2,D18P186001R2,58,2.440946125,,,,D18P186001R2T58,
D18P186001,R2,D18P186001R2,59,5.15748325,,ABBA,,D18P186001R2T59,
D18P186001,R2,D18P186001R2,60,1.653544125,,ABBA,,D18P186001R2T60,
D18P186001,R2,D18P186001R2,61,,,BEPA,,D18P186001R2T61,
D18P186001,R2,D18P186001R2,62,0.944882438,,BEPA,,D18P186001R2T62,
D18P186001,R2,D18P186001R2,63,1.102362781,,BEPA,,D18P186001R2T63,
D18P186001,R2,D18P186001R2,64,1.811024562,,ABBA,,D18P186001R2T64,
D18P186001,R2,D18P186001R2,65,0.708661781,,BEPA,,D18P186001R2T65,
D18P186001,R2,D18P186001R2,66,7.12598825,,ABBA,,D18P186001R2T66,
D18P186001,R2,D18P186001R2,67,7.51968925,,ABBA,36,D18P186001R2T67,
D18P186001,R2,D18P186001R2,68,1.3779535,,ABBA,,D18P186001R2T68,
D18P186001,R2,D18P186001R2,69,6.14173575,,ABBA,,D18P186001R2T69,
D18P186001,R2,D18P186001R2,70,8.89764275,,ABBA,30,D18P186001R2T70,
D18P186001,R2,D18P186001R2,71,0.787402,,BEPA,,D18P186001R2T71,
D18P186001,R2,D18P186001R2,72,5.826774875,,ABBA,27,D18P186001R2T72,
D18P186001,R2,D18P186001R2,73,0.354330891,,BEPA,,D18P186001R2T73,
D18P186001,R2,D18P186001R2,74,0.511811281,,BEPA,,D18P186001R2T74,
D18P186001,R2,D18P186001R2,75,0.433071109,,BEPA,,D18P186001R2T75,
D18P186001,R2,D18P186001R2,76,0.551181391,,BEPA,,D18P186001R2T76,
D18P186001,R2,D18P186001R2,77,1.062992719,,ABBA,,D18P186001R2T77,
D18P186001,R2,D18P186001R2,78,0.511811281,,BEPA,,D18P186001R2T78,
D18P186001,R2,D18P186001R2,79,5.826774875,,ABBA,,D18P186001R2T79,
D18P186001,R2,D18P186001R2,80,8.50394175,,ABBA,,D18P186001R2T80,
D18P186001,R2,D18P186001R2,81,,,ABBA,,D18P186001R2T81,
D18P186001,R2,D18P186001R2,82,0.157480402,,BEPA,,D18P186001R2T82,
D18P186001,R2,D18P186001R2,83,0.118110305,,BEPA,,D18P186001R2T83,
D18P186001,R2,D18P186001R2,84,,,BEPA,,D18P186001R2T84,
D18P186001,R2,D18P186001R2,85,,,BEPA,,D18P186001R2T85,
D18P186001,R2,D18P186001R2,86,0.118110305,,BEPA,,D18P186001R2T86,
D18P186001,R2,D18P186001R2,87,,,BEPA,,D18P186001R2T87,
D18P186001,R2,D18P186001R2,88,0.314960805,,BEPA,,D18P186001R2T88,
D18P186001,R2,D18P186001R2,89,,,BEPA,,D18P186001R2T89,
D18P186001,R2,D18P186001R2,90,0.433071109,,BEPA,,D18P186001R2T90,
D18P186001,R2,D18P186001R2,91,0.078740201,,BEPA,,D18P186001R2T91,
D18P186001,R2,D18P186001R2,92,0.708661781,,ABBA,,D18P186001R2T92,
D18P186001,R2,D18P186001R2,93,0.393701,,ABBA,,D18P186001R2T93,
D18P186001,R2,D18P186001R2,94,0.472441219,,ABBA,,D18P186001R2T94,
D18P186001,R2,D18P186001R2,95,,,BEPA,,D18P186001R2T95,
D18P186001,R2,D18P186001R2,96,0.393701,,BEPA,,D18P186001R2T96,
D18P186001,R2,D18P186001R2,97,0.393701,,BEPA,,D18P186001R2T97,
D18P186001,R2,D18P186001R2,98,0.393701,,BEPA,,D18P186001R2T98,
D18P186001,R2,D18P186001R2,99,0.472441219,,BEPA,,D18P186001R2T99,
D18P186001,R2,D18P186001R2,100,0.433071109,,BEPA,,D18P186001R2T100,
D18P186001,R3,D18P186001R3,1,6.92913775,,ABBA,,D18P186001R3T1,
D18P186001,R3,D18P186001R3,2,1.456693719,,BEPA,,D18P186001R3T2,
D18P186001,R3,D18P186001R3,3,,,BEPA,,D18P186001R3T3,
D18P186001,R3,D18P186001R3,4,,,BEPA,,D18P186001R3T4,
D18P186001,R3,D18P186001R3,5,,,BEPA,,D18P186001R3T5,
D18P186001,R3,D18P186001R3,6,4.48819125,,ABBA,37,D18P186001R3T6,
D18P186001,R3,D18P186001R3,7,3.858269875,,ABBA,,D18P186001R3T7,
D18P186001,R3,D18P186001R3,8,,,,,D18P186001R3T8,
D18P186001,R3,D18P186001R3,9,2.716536938,,,,D18P186001R3T9,
D18P186001,R3,D18P186001R3,10,,,,,D18P186001R3T10,
D18P186001,R3,D18P186001R3,11,1.535433938,,BEPA,,D18P186001R3T11,
D18P186001,R3,D18P186001R3,12,5.590554125,,ABBA,,D18P186001R3T12,
D18P186001,R3,D18P186001R3,13,4.37008125,,ABBA,,D18P186001R3T13,
D18P186001,R3,D18P186001R3,14,1.338583438,,ABBA,,D18P186001R3T14,
D18P186001,R3,D18P186001R3,15,5.984255125,,ABBA,39,D18P186001R3T15,
D18P186001,R3,D18P186001R3,16,4.88189225,,ABBA,39,D18P186001R3T16,
D18P186001,R3,D18P186001R3,17,1.338583438,,ABBA,,D18P186001R3T17,
D18P186001,R3,D18P186001R3,18,1.141732938,,BEPA,,D18P186001R3T18,
D18P186001,R3,D18P186001R3,19,,,BEPA,,D18P186001R3T19,
D18P186001,R3,D18P186001R3,20,0.669291719,,BEPA,,D18P186001R3T20,
D18P186001,R3,D18P186001R3,21,3.070867875,,ABBA,33,D18P186001R3T21,
D18P186001,R3,D18P186001R3,22,,,BEPA,,D18P186001R3T22,
D18P186001,R3,D18P186001R3,23,1.259843219,,ABBA,,D18P186001R3T23,
D18P186001,R3,D18P186001R3,24,5.905515,,ABBA,36,D18P186001R3T24,
D18P186001,R3,D18P186001R3,25,5.433073875,,ABBA,,D18P186001R3T25,
D18P186001,R3,D18P186001R3,26,,,BEPA,,D18P186001R3T26,
D18P186001,R3,D18P186001R3,27,1.023622562,,BEPA,,D18P186001R3T27,
D18P186001,R3,D18P186001R3,28,4.96063275,,ABBA,,D18P186001R3T28,
D18P186001,R3,D18P186001R3,29,,,BEPA,,D18P186001R3T29,
D18P186001,R3,D18P186001R3,30,0.511811281,,BEPA,,D18P186001R3T30,
D18P186001,R3,D18P186001R3,31,6.6141765,,ABBA,,D18P186001R3T31,
D18P186001,R3,D18P186001R3,32,6.06299525,,ABBA,,D18P186001R3T32,
D18P186001,R3,D18P186001R3,33,,,,,D18P186001R3T33,
D18P186001,R3,D18P186001R3,34,0.9842525,,,,D18P186001R3T34,
D18P186001,R3,D18P186001R3,35,,,,,D18P186001R3T35,
D18P186001,R3,D18P186001R3,36,1.220473062,,,,D18P186001R3T36,
D18P186001,R3,D18P186001R3,37,5.236223375,,ABBA,41,D18P186001R3T37,
D18P186001,R3,D18P186001R3,38,6.4960665,,ABBA,,D18P186001R3T38,
真正的答案
基本上，您编写的代码就像您用其他语言编写过很多但从未使用过的代码 R。由于所有嵌套循环，很难准确理解发生了什么，而且我没有 运行 你的代码（请参阅 here 了解如何可重复地包含数据）。
我认为虽然可以删除所有循环，例如用类似的东西替换第一个条件：
FVS_TreeInit$GD[
FVS_TreeInit$Plot_ID == "R8" &
FVS_TreeInit$StandPlotTree_ID == paste(FVS_TreeInit$Stand_ID,"R7T",FVS_TreeInit$Tree_ID,sep="")
] <- FVS_TreeInit$DBH
如果您使用 data.table（此处），它将是 significantly quicker，因为它的实现速度如此之快。
便宜又简单的答案
有一种既便宜又简单的方法可以在不重构所有代码的情况下加快您现有的速度，但它仍然会很慢。
问题是您一直在使用 df[i, "columnName"] 查找数据。这是一种非常缓慢的方法。看这个例子。
创建两个随机的大（10e6 行）字符串数据框：
# Random data function borrowed from: https://www.r-bloggers.com/2017/03/fast-data-lookups-in-r-dplyr-vs-data-table/
random_string_column <- function(n) {
stringi::stri_rand_strings(n = n, length = 8)
}
random_data_frame <- function(n) tibble(
col1 = random_string_column(n),
col2 = random_string_column(n)
)
df <- random_data_frame(10^6)
df2 <- random_data_frame(10^6)
head(df)
输出（df2会类似）：
# A tibble: 6 x 2
col1 col2
<chr> <chr>
1 NykzxkNO qGusGkMa
2 mgivTVky xlCtwEba
3 zar6FbC4 dOYeQPfg
4 89fKLJFv RPsP2CZc
5 V2FH3Smt ObToBJFC
6 gXa0j5XZ QYKK6hD2
现在让我们按照您的方式进行 1000 次随机查找：
rand_indeces <- sample(c(1:nrow(df), 1000))
system.time(
for(i in rand_indeces) {
if(df[i, "col2"]=="aaaaa") { #it never will
break
}
}
)
输出：
user system elapsed
138.58 0.23 140.72
两分钟多了。现在让我们看看另一个数据框（新的 df 所以我们可以看到它不是因为缓存），用 df2$col2[i] == "aaaaa".
df[i, "col2"]=="aaaaa" 语法
system.time(
for(i in rand_indeces) {
if(df2$col2[i] == "aaaaa") { #it never will
break
}
}
)
输出：
user system elapsed
4.70 0.00 4.75
这应该可以为您节省不少时间。但是，从根本上说，在 R 中以这种方式进行查找很慢（参见 here and
如果您必须进行单独的查找，您可以考虑创建一个散列 table，或者通过第一个 link 中的 environments 或者通过一些较新的方法 here.我认为 collections 包看起来很有趣 - 还没有使用它。
但是，到目前为止，最好的办法是矢量化您的代码，如本回复顶部所述。
（扰流板：跳转到 Double Merge 部分以获得比嵌套循环快 150 倍的结果。我希望这种速度改进随着数据的增加而变得更好。）
中级改进
首先，您的内部循环都可以从 for 循环中减少，从而变得更快，也许会改变
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
for (f in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[f,"StandPlotTree_ID"] == find){
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
break
}
}
到
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
f <- FVS_TreeInit[,"StandPlotTree_ID"] == find
FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
值得注意的是，我们已经从迭代 for 循环更改为 向量化 比较，这在 R 中往往要快得多。但是，应该注意我们对几个 "R*" 做同样的事情，所以我们可以大大减少你的外层循环；其中大部分是代码高尔夫（减少字符数来做同样的事情），但总的来说不要重复你自己：如果你有相同的（相同的）代码多个地方，你增加了改变一个而不是所有其他地方的机会。
在那个主题中，到目前为止，我可以通过使用这个得到相同的结果：
# loop through FVS_TreeInit and add previous diameter increment identifier to current row
for (i in seq_len(nrow(FVS_TreeInit))){
if (FVS_TreeInit[i,"Plot_ID"] == "R8"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R7T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R7"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R6T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R6"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R5T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R5"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R4T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R4"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R3T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R3"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R2T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R2"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R1T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else if (FVS_TreeInit[i,"Plot_ID"] == "R1"){
find <- paste(FVS_TreeInit[i,"Stand_ID"],"R0T",FVS_TreeInit[i,"Tree_ID"],sep="")
} else {
find <- NA
}
FVS_TreeInit[i,"GD"] <- NA
if (!is.na(find)) {
f <- FVS_TreeInit[,"StandPlotTree_ID"] == find
if (any(f)) FVS_TreeInit[i,"GD"] <- FVS_TreeInit[f,"DBH"]
}
if (i%%1000 == 0){
print(paste(i," rows iterated...",sep=""))
}
}
好的，那更短了，缩小到真正重要的部分。值得注意的是（对我而言），我们在这里根据某些特定值有效地将数据“连接”在一起。我认为我们可以通过更简单的步骤（合并）完成所有这些工作。
双重合并
这实际上只是根据三个 ID：Stand_ID、Tree_ID 和 Plot_ID 将数据自身连接回去，其中后者是一个时间组件（因此需要为自加入进行调整）。
基于您的 R* 值反映时间的想法，我认为我们可以安全地假设 R(n) 应该分配给 R(n+1)，所以让我们以编程方式创建这个中间框架：
prevPlot <- data.frame(prev_Plot_ID = unique(FVS_TreeInit$Plot_ID))
prevPlot$Plot_ID <- paste0("R", as.integer(gsub("R", "", prevPlot$prev_Plot_ID))-1)
prevPlot <- subset(prevPlot, prev_Plot_ID != "R0")
prevPlot
# prev_Plot_ID Plot_ID
# 2 R1 R0
# 3 R2 R1
# 4 R3 R2
我们将 merge/join 它放到原来的 FVS_Tree_Init 上以识别“先前的 Plot_ID”，然后合并回自身以引入新值。
prevValues <- subset(merge(FVS_TreeInit, prevPlot, by = "Plot_ID", all.x = TRUE),
select = c(Stand_ID, prev_Plot_ID, Tree_ID, DBH))
names(prevValues)[4] <- "GD"
head(prevValues)
# Stand_ID prev_Plot_ID Tree_ID GD
# 1 D18P186001 R1 39 3.9370100
# 2 D18P186001 R1 40 4.2519709
# 3 D18P186001 R1 1 4.7637823
# 4 D18P186001 R1 2 1.2598432
# 5 D18P186001 R1 3 0.2755907
# 6 D18P186001 R1 4 0.6299216
new_FVS_TreeInit <- merge(
FVS_TreeInit, prevValues,
by.x = c("Stand_ID", "Plot_ID", "Tree_ID"),
by.y = c("Stand_ID", "prev_Plot_ID", "Tree_ID"),
all.x = TRUE)
identical(expected, new_FVS_TreeInit[names(expected)])
# [1] TRUE
其中 expected 是嵌套 for 循环的输出，new_FVS_TreeInit[names(expected)] 只是确保所有列顺序相同的一个步骤（与 identical，数据处理不需要）。
仅供参考，这突出了您使用连接字符串作为单数键的矛盾（一种相当 Excel- 进行连接的方法......它确实有效）前提是 merging/joining 一个或多个键上的数据。一旦您放弃单键 Excel 方法并认识到 一个或多个 也同样有效——并且更容易阅读和维护——我认为转变观念合并可以改变编程。有关联接“演算”的更多讨论，请参阅 How to join (merge) data frames (inner, outer, left, right) and What's the difference between INNER JOIN, LEFT JOIN, RIGHT JOIN and FULL JOIN?。
基准
使用此示例数据，执行速度大约 150 倍。
bench::mark(
forloop = {
# your code
},
doublemerge = {
# my code
}
)
# Warning: Some expressions had a GC in every iteration; so filtering is disabled.
# # A tibble: 2 x 13
# expression min median `itr/sec` mem_alloc `gc/sec` n_itr n_gc total_time result memory time gc
# <bch:expr> <bch:tm> <bch:tm> <dbl> <bch:byt> <dbl> <int> <dbl> <bch:tm> <list> <list> <list> <list>
# 1 forloop 272.4ms 284.86ms 3.51 656KB 7.02 2 4 570ms <df [318 x 11]> <Rprofmem ~ <bench_t~ <tibble ~
# 2 doublemerge 1.4ms 1.55ms 528. 346KB 1.99 265 1 502ms <df [318 x 11]> <Rprofmem ~ <bench_t~ <tibble ~
单程data.table解决方案：
这可以在依次按 Stand_ID、Tree_ID、Plot_ID 排序后一次性完成（在我的机器上约 6 毫秒，包括读取数据）。
首先，我将示例数据字符串复制到对象 data。
library(data.table)
dt <- fread(input = data) # read into a data.table
setorder(dt[, `:=`(Plot_ID = as.integer(substr(Plot_ID, 2, 2)), DG = as.numeric(DG))], # convert Plot_ID to an integer and DG to numeric (from logical)
# order by Stand_ID, Tree_ID, Plot_ID
Stand_ID, Tree_ID, Plot_ID)[
# create a temporary column indicating the rows that will pull DBH from the previous row
, pull := c(FALSE, diff(Plot_ID) == 1L) & Stand_ID == shift(Stand_ID) & Tree_ID == shift(Tree_ID)][
# pull DBH from the previous row where appropriate
pull == TRUE, DG := shift(dt$DBH)[dt$pull]][,
# clean up
`:=`(pull = NULL, Plot_ID = paste0("R", Plot_ID))]
我相信这可以用更少的代码以更简单的方式重构。看起来您正在对数据进行连接，其中查找键是当前 Plot_ID 的修改版本，其中“R#”应链接到“R#-1T”。
在 R 中执行此操作的更惯用的方法是为所有行创建一次适当的键，然后根据该查找键进行连接。这将比您的循环方法快几个数量级，因为它只需要编译查找代码一次而不是 650,000 次，并且连接针对速度进行了优化。请参阅此页面以深入解释为什么向量化在 R 中有如此大的帮助：https://www.noamross.net/archives/2014-04-16-vectorization-in-r-why/
我在下面使用 dplyr 执行此操作，但该方法在 base r 中的工作方式非常相似。
library(dplyr)
FVS_TreeInit %>%
# create key for all rows
mutate(prior_key = paste0(Stand_ID, "R",
parse_number(Plot_ID)-1, "T", # will be nonsense R-1T when starts from R0, that's ok,
Tree_ID)) %>% # it just won't replace GD there
# Use that key to join to a version of the data with just StandPlotTree_ID and
# DBH, which I rename as GD. Match `prior_key` with `StandPlotTree_ID` there.
left_join(FVS_TreeInit %>% transmute(StandPlotTree_ID, GD = DBH),
by = c("prior_key" = "StandPlotTree_ID"))
基准
您的原始代码在提供的示例数据上 运行 花费了 3.7 秒。我建议的更换用了 0.027 秒，大约是更换速度的 140 倍。我预计随着数据的增加，时差会变得更大。