groupMeans( x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupSums( x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupVars( x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupStdevs(x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupMaxs(x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupMins(x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupProds(x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupRanges(x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupAnys(x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) groupAlls(x, group=NULL, na.rm=F, weights=NULL, repeats=F, ...) subtractMeans(x, group=NULL, na.rm=F, weights=NULL)
numRows(x)
. Unique values of
this vector determine groups (sets of rows).
Statistics are
calculated for each group separately, and for each column.
The default corresponds to a single group.
FALSE
, missing values (
NA
) in the input result in
missing values in corresponding elements of the output.
If
TRUE
then missing values are omitted from calculations.
numRows(x)
. If present then weighted
statistics are calculated.
T
,
F
,
0
,
1
, or
2
. See "VALUE" below.
groupVars
and
groupStdevs
accept arguments
unbiased
and
SumSquares
;
see
.
repeats=F
or
0
then the
"group"
functions return one
statistic value per stratum per column of
x
. Thus if
x
is an
array, the first dimension of the result is equal to the number of
groups, and all other dimensions are the same as for
x
. If
x
is a
vector, the functions return a vector (if more than one group) or
scalar (if one group). The names or row names of the result
correspond to the sorted unique values of
group
.
The exception is
groupRanges
, which returns two values per stratum.
In this case the first dimension of the output corresponds to the two
values, the second dimension to groups, and other dimensions as for
x
.
If
repeats=T
or
1
the
"group"
functions return an object with the same
dimensions as
x
: each entry is the value of the statistic for the
stratum containing the corresponding entry from
x
.
(For
groupRanges
a dimension of 2 is prepended.)
If
repeats=2
,
return a list with both the repeated and non-repeated objects.
The list has names that match the statistic --
"means"
and
"repeat.means"
, for example.
subtractMeans
returns a vector or array the same size as
x
, with
value equal to
x
minus
"repeat.means"
.
# Center Mileage data about the averages for each type of vehicle subtractMeans(fuel.frame$Mileage,group=fuel.frame$Type) # same, randomly weighted means subtractMeans(fuel.frame$Mileage,group=fuel.frame$Type, weights=runif(60)) # means and standard deviations for Types Small and Sporty groupMeans(fuel.frame$Mileage, group=fuel.frame$Type)[c("Small", "Sporty")] groupStdevs(fuel.frame$Mileage, group=fuel.frame$Type)[c("Small", "Sporty")]