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
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
|
% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/createDataPartition.R
\name{createDataPartition}
\alias{createDataPartition}
\alias{createFolds}
\alias{createMultiFolds}
\alias{createResample}
\alias{createTimeSlices}
\title{Data Splitting functions}
\usage{
createDataPartition(y, times = 1, p = 0.5, list = TRUE, groups = min(5,
length(y)))
createFolds(y, k = 10, list = TRUE, returnTrain = FALSE)
createTimeSlices(y, initialWindow, horizon = 1, fixedWindow = TRUE,
skip = 0)
}
\arguments{
\item{y}{a vector of outcomes. For \code{createTimeSlices}, these should be
in chronological order.}
\item{times}{the number of partitions to create}
\item{p}{the percentage of data that goes to training}
\item{list}{logical - should the results be in a list (\code{TRUE}) or a
matrix with the number of rows equal to \code{floor(p * length(y))} and
\code{times} columns.}
\item{groups}{for numeric \code{y}, the number of breaks in the quantiles
(see below)}
\item{k}{an integer for the number of folds.}
\item{returnTrain}{a logical. When true, the values returned are the sample
positions corresponding to the data used during training. This argument only
works in conjunction with \code{list = TRUE}}
\item{initialWindow}{The initial number of consecutive values in each
training set sample}
\item{horizon}{The number of consecutive values in test set sample}
\item{fixedWindow}{A logical: if \code{FALSE}, the training set always start
at the first sample.}
\item{skip}{An integer specifying how many (if any) resamples to skip to
thin the total amount.}
}
\value{
A list or matrix of row position integers corresponding to the
training data
}
\description{
A series of test/training partitions are created using
\code{createDataPartition} while \code{createResample} creates one or more
bootstrap samples. \code{createFolds} splits the data into \code{k} groups
while \code{createTimeSlices} creates cross-validation sample information to
be used with time series data.
}
\details{
For bootstrap samples, simple random sampling is used.
For other data splitting, the random sampling is done within the levels of
\code{y} when \code{y} is a factor in an attempt to balance the class
distributions within the splits.
For numeric \code{y}, the sample is split into groups sections based on
percentiles and sampling is done within these subgroups. For
\code{createDataPartition}, the number of percentiles is set via the
\code{groups} argument. For \code{createFolds} and \code{createMultiFolds},
the number of groups is set dynamically based on the sample size and
\code{k}. For smaller samples sizes, these two functions may not do
stratified splitting and, at most, will split the data into quartiles.
Also, for \code{createDataPartition}, very small class sizes (<= 3) the
classes may not show up in both the training and test data
For multiple k-fold cross-validation, completely independent folds are
created. The names of the list objects will denote the fold membership
using the pattern "Foldi.Repj" meaning the ith section (of k) of the jth
cross-validation set (of \code{times}). Note that this function calls
\code{createFolds} with \code{list = TRUE} and \code{returnTrain = TRUE}.
Hyndman and Athanasopoulos (2013)) discuss rolling forecasting origin
techniques that move the training and test sets in time.
\code{createTimeSlices} can create the indices for this type of splitting.
}
\examples{
data(oil)
createDataPartition(oilType, 2)
x <- rgamma(50, 3, .5)
inA <- createDataPartition(x, list = FALSE)
plot(density(x[inA]))
rug(x[inA])
points(density(x[-inA]), type = "l", col = 4)
rug(x[-inA], col = 4)
createResample(oilType, 2)
createFolds(oilType, 10)
createFolds(oilType, 5, FALSE)
createFolds(rnorm(21))
createTimeSlices(1:9, 5, 1, fixedWindow = FALSE)
createTimeSlices(1:9, 5, 1, fixedWindow = TRUE)
createTimeSlices(1:9, 5, 3, fixedWindow = TRUE)
createTimeSlices(1:9, 5, 3, fixedWindow = FALSE)
createTimeSlices(1:15, 5, 3)
createTimeSlices(1:15, 5, 3, skip = 2)
createTimeSlices(1:15, 5, 3, skip = 3)
}
\author{
Max Kuhn, \code{createTimeSlices} by Tony Cooper
}
\references{
\url{http://topepo.github.io/caret/splitting.html}
Hyndman and Athanasopoulos (2013), Forecasting: principles and practice.
\url{https://www.otexts.org/fpp}
}
\keyword{utilities}
|
