以下に定義されている問題の解決策をいただければ幸いです。とても難しいことだと思います。
2 つの data.frames を結合t1しt2、merge(). 「testing」と名付けた結果の data.frame で、一意でない行のエントリt1を「NA」に置き換えて、最も近い距離を持つ一意の行のみが残るようにしt2ます。条件は次のとおりです。
min(sqrt((xCor.y - xCor.x)^2 + (yCor.y - yCor.x)^2))
# scroll to end for result I am looking for
これは最大 1 GB のデータ セットを対象としているため、すべてのデータをループ処理しないようにする必要があります。
t1<- data.frame(trackLabel = c(1, 2, 3, 4, 4, 5, 5, 7, 7, 7),
objNumber = 1:10,
parentObjNumber = rep(0, 10),
time = rep(1,10),
xCor = runif(10),
yCor = runif(10))
t2<- data.frame(trackLabel = c(1, 2, 2, 4, 4, 4, 6, 7, 7, 7, 7),
objNumber = 11:21,
parentObjNumber = c(1, 2, 2, 4, 4, 4, 7, 8, 9, 9, 9),
time = rep(2,11),
xCor = runif(11),
yCor = runif(11))
testing<- merge(t1, t2,
by.x = c("trackLabel", "objNumber"),
by.y = c("trackLabel", "parentObjNumber"),
all = TRUE,
incomparables = NA)
#Warning message:
#In merge.data.frame(t1, t2, by.x = c("trackLabel", "objNumber"), :
# column name ‘objNumber’ is duplicated in the result
ind<-colnames(testing)=="objNumber"
colnames(testing)[min(which(ind == TRUE )) ] <- paste("objNumber", 1, sep = "")
> t1
trackLabel objNumber parentObjNumber time xCor yCor
1 1 1 0 1 0.25852366 0.360631607
2 2 2 0 1 0.69987607 0.048360258
3 3 3 0 1 0.23047883 0.414221880
4 4 4 0 1 0.58169548 0.718223111
5 4 5 0 1 0.61419336 0.435153774
6 5 6 0 1 0.50028765 0.735970291
7 5 7 0 1 0.41380332 0.097256739
8 7 8 0 1 0.57563080 0.828142024
9 7 9 0 1 0.39512092 0.728903233
10 7 10 0 1 0.16675690 0.284307824
> t2
trackLabel objNumber parentObjNumber time xCor yCor
1 1 11 1 2 0.473735625 0.454637752
2 2 12 2 2 0.623971860 0.517089522
3 2 13 2 2 0.470885840 0.703872484
4 4 14 4 2 0.188280842 0.678683831
5 4 15 4 2 0.198772198 0.160836676
6 4 16 4 2 0.251950005 0.958747183
7 6 17 7 2 0.545521560 0.005505346
8 7 18 8 2 0.477450908 0.819060935
9 7 19 9 2 0.509430458 0.997968108
10 7 20 9 2 0.027918865 0.138014769
11 7 21 9 2 0.568532497 0.911921770
> testing
trackLabel objNumber1 parentObjNumber time.x xCor.x yCor.x objNumber time.y xCor.y yCor.y
1 1 1 0 1 0.25852366 0.360631607 11 2 0.473735625 0.454637752
2 2 2 0 1 0.69987607 0.048360258 12 2 0.623971860 0.517089522
3 2 2 0 1 0.69987607 0.048360258 13 2 0.470885840 0.703872484
4 3 3 0 1 0.23047883 0.414221880 NA NA NA NA
5 4 4 0 1 0.58169548 0.718223111 14 2 0.188280842 0.678683831
6 4 4 0 1 0.58169548 0.718223111 15 2 0.198772198 0.160836676
7 4 4 0 1 0.58169548 0.718223111 16 2 0.251950005 0.958747183
8 4 5 0 1 0.61419336 0.435153774 NA NA NA NA
9 5 6 0 1 0.50028765 0.735970291 NA NA NA NA
10 5 7 0 1 0.41380332 0.097256739 NA NA NA NA
11 6 7 NA NA NA NA 17 2 0.545521560 0.005505346
12 7 8 0 1 0.57563080 0.828142024 18 2 0.477450908 0.819060935
13 7 9 0 1 0.39512092 0.728903233 19 2 0.509430458 0.997968108
14 7 9 0 1 0.39512092 0.728903233 20 2 0.027918865 0.138014769
15 7 9 0 1 0.39512092 0.728903233 21 2 0.568532497 0.911921770
16 7 10 0 1 0.16675690 0.284307824 NA NA NA NA
# and here is what I want to achieve:
> testing[c(3, 6, 7, 13, 14 ), 1:6] <-NA
> testing
trackLabel objNumber1 parentObjNumber time.x xCor.x yCor.x objNumber time.y xCor.y yCor.y
1 1 1 0 1 0.25852366 0.360631607 11 2 0.473735625 0.454637752
2 2 2 0 1 0.69987607 0.048360258 12 2 0.623971860 0.517089522
3 NA NA NA NA NA NA 13 2 0.470885840 0.703872484
4 3 3 0 1 0.23047883 0.414221880 NA NA NA NA
5 4 4 0 1 0.58169548 0.718223111 14 2 0.188280842 0.678683831
6 NA NA NA NA NA NA 15 2 0.198772198 0.160836676
7 NA NA NA NA NA NA 16 2 0.251950005 0.958747183
8 4 5 0 1 0.61419336 0.435153774 NA NA NA NA
9 5 6 0 1 0.50028765 0.735970291 NA NA NA NA
10 5 7 0 1 0.41380332 0.097256739 NA NA NA NA
11 6 7 NA NA NA NA 17 2 0.545521560 0.005505346
12 7 8 0 1 0.57563080 0.828142024 18 2 0.477450908 0.819060935
13 NA NA NA NA NA NA 19 2 0.509430458 0.997968108
14 NA NA NA NA NA NA 20 2 0.027918865 0.138014769
15 7 9 0 1 0.39512092 0.728903233 21 2 0.568532497 0.911921770
16 7 10 0 1 0.16675690 0.284307824 NA NA NA NA