memisc (version 0.99.22)

Reshape: Reshape data frames or data sets

Description

Reshape is a convenience wrapper around reshape with a somewhat simpler syntax.

Usage

Reshape(data,…,id,within_id,drop,direction)

Arguments

data

a data frame or data set to be reshaped.

…

Further arguments that specify the variables in long and in wide format as well as the time variable. The name tags of the arguments given here specify variable names in long format, the arguments themselves specify the variables in wide format (or observations in long vormat) and the variable of the "time" variable. The time variable is usually the last of these arguments. An "automatic" time variable can be specified if only a single argument in … is given.

id

a variable name or a concatenation of variable names (either as character strings or as unquoted symbols), that identify individual units. Defaults to "id" or the id variable specified in the "reshapeLong" attribute of the data argument. Needed only if the data are reshaped from long to wide format.

within_id

an optional variable name (either as character string or as unquoted symbol), that identifies individual observations on units. Relevant only if the data are reshaped from long to wide format.

drop

a variable name or a concatenation of variable names (either as character strings or as unquoted symbols), thast specifies the variables to be dropped before reshaping.

direction

a character string, should be either equal "long" or "wide".

Examples

# NOT RUN {
example.data.wide <- data.frame(
    v  = c(35,42),
    x1 = c(1.1,2.1),
    x2 = c(1.2,2.2),
    x3 = c(1.3,2.3),
    x4 = c(1.4,2.4),
    y1 = c(2.5,3.5),
    y2 = c(2.7,3.7),
    y3 = c(2.9,3.9))
example.data.wide

# The following two calls are equivalent:
example.data.long <- Reshape(data=example.data.wide,
                             x=c(x1,x2,x3,x4),
                             # N.B. it is possible to
                             # specify 'empty' i.e. missing
                             # measurements
                             y=c(y1,y2,y3,),
                             t=1:4,
                             direction="long")

example.data.long <- Reshape(data=example.data.wide,
                             list(
                                 x=c(x1,x2,x3,x4),
                                 # N.B. it is possible to
                                 # specify 'empty' i.e. missing
                                 # measurements
                                 y=c(y1,y2,y3,)
                             ),
                             t=1:4,
                             direction="long")

example.data.long

# Since the data frame contains an "reshapeLong" attribute
# an id variable is already specified and part of the data
# frame.
example.data.wide <- Reshape(data=example.data.long,
                             x=c(x1,x2,x3,x4),
                             y=c(y1,y2,y3,),
                             t=1:4,
                             direction="wide")
example.data.wide

# Here we examine the case where no "reshapeLong" attribute
# is present:
example.data.wide <- Reshape(data=example.data.long,
                             x=c(x1,x2,x3,x4),
                             y=c(y1,y2,y3,),
                             t=1:4,
                             id=v,
                             direction="wide")
example.data.wide


# Here, an "automatic" time variable is created. This works
# only if there is a single argument other than the data=
# and direction= arguments
example.data.long <- Reshape(data=example.data.wide,
                             list(
                                 x=c(x1,x2,x3,x4),
                                 y=c(y1,y2,y3,)
                             ),
                             direction="long")

example.data.long

example.data.wide <- Reshape(data=example.data.long,
                             list(
                                 x=c(x1,x2,x3,x4),
                                 y=c(y1,y2,y3,)
                             ),
                             direction="wide")
example.data.wide
# }