1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
|
instantiateResampleInstance = function(desc, size, task) {
UseMethod("instantiateResampleInstance")
}
instantiateResampleInstance.HoldoutDesc = function(desc, size, task = NULL) {
inds = sample(size, size * desc$split)
makeResampleInstanceInternal(desc, size, train.inds = list(inds))
}
instantiateResampleInstance.CVDesc = function(desc, size, task = NULL) {
if (desc$iters > size)
stopf("Cannot use more folds (%i) than size (%i)!", desc$iters, size)
test.inds = chunk(seq_len(size), shuffle = TRUE, n.chunks = desc$iters)
makeResampleInstanceInternal(desc, size, test.inds = test.inds)
}
instantiateResampleInstance.SpCVDesc = function(desc, size, task = NULL) {
if (is.null(task)) {
stopf("Please provide a task.")
}
if (is.null(task$coordinates)) {
stopf("Please provide suitable coordinates for SpCV. See ?Task for help.")
}
# perform kmeans clustering
inds = kmeans(task$coordinates, centers = desc$iters)
inds = factor(inds$cluster)
# uses resulting factor levels from kmeans clustering to set up a list of
# length x (x = folds) with row indices of the data referring to which fold
# each observations is assigned to
test.inds = lapply(levels(inds), function(x, spl)
which(spl == x), spl = inds)
makeResampleInstanceInternal(desc, size, test.inds = test.inds)
}
instantiateResampleInstance.LOODesc = function(desc, size, task = NULL) {
desc$iters = size
makeResampleInstanceInternal(desc, size, test.inds = as.list(seq_len(size)))
}
instantiateResampleInstance.SubsampleDesc = function(desc, size, task = NULL) {
inds = lapply(seq_len(desc$iters), function(x) sample(size, size * desc$split))
makeResampleInstanceInternal(desc, size, train.inds = inds)
}
instantiateResampleInstance.BootstrapDesc = function(desc, size, task = NULL) {
inds = lapply(seq_len(desc$iters), function(x) sample(size, size, replace = TRUE))
makeResampleInstanceInternal(desc, size, train.inds = inds)
}
instantiateResampleInstance.RepCVDesc = function(desc, size, task = NULL) {
folds = desc$iters / desc$reps
d = makeResampleDesc("CV", iters = folds)
i = replicate(desc$reps, makeResampleInstance(d, size = size), simplify = FALSE)
train.inds = Reduce(c, lapply(i, function(j) j$train.inds))
test.inds = Reduce(c, lapply(i, function(j) j$test.inds))
g = as.factor(rep(seq_len(desc$reps), each = folds))
makeResampleInstanceInternal(desc, size, train.inds = train.inds, test.inds = test.inds, group = g)
}
instantiateResampleInstance.SpRepCVDesc = function(desc, size, task = NULL) {
folds = desc$iters / desc$reps
d = makeResampleDesc("SpCV", iters = folds)
i = replicate(desc$reps, makeResampleInstance(d, task = task), simplify = FALSE)
train.inds = Reduce(c, lapply(i, function(j) j$train.inds))
test.inds = Reduce(c, lapply(i, function(j) j$test.inds))
g = as.factor(rep(seq_len(desc$reps), each = folds))
makeResampleInstanceInternal(desc, size, train.inds = train.inds, test.inds = test.inds, group = g)
}
instantiateResampleInstance.FixedWindowCVDesc = function(desc, size, task = NULL, coords) {
makeResamplingWindow(desc, size, task, coords, "FixedWindowCV")
}
instantiateResampleInstance.GrowingWindowCVDesc = function(desc, size, task = NULL, coords) {
makeResamplingWindow(desc, size, task, coords, "GrowingWindowCV")
}
|
