egodata
Objectsas.egodata
is a generic function to construct
egodata
objects from a variety of sources.
egodata
function is the standard constructor, taking two data
frames. For other methods for this class, see the Miscellaneous Methods
section.
egodata(egos, alters, egoWt = 1, ..., egoIDcol = "egoID")as.egodata(object, ..., egoIDcol = "egoID")
# S3 method for data.frame
as.egodata(object, alters, egoWt = 1, ...,
egoIDcol = "egoID", alterIcol = "alterInd", alterIDcol = "alterID")
The data.frame
containing at least the column
named in egoIDcol
, whose values do not have to be unique, and
not every ego has to be represented. Other columns contain information
about the alters.
For the data.frame
method in which the object
argument also contains alter information in ‘wide’ format, a
list
with the following information:
columns
A character, integer, or logical vector identifying which columns contain the alters' information.
count
The name of the column containing the number of alters nominated by that ego.
name.sep
A one-character string or an empty string
(defaulting to "."
) specifying the character, if any,
used to separate alter attribute name from alter's index within
the ego. If an empty string (""
), attribute name is
assumed to be made of letters, with any numbers being the alter
index.
egoIDcol
, whose values do not have to be unique, and not every ego
must be represented. Other columns contain information about the alters.
A vector of the same length as number of rows in egos
or
object
, containing the relative sampling weight of each ego.
Additional arguments; currently unused.
Name of the column in the ego table containing the unique ego identifier.
The object from which the egocentric data should be
constructed. For the data.frame
methods and
egodata
itself, a data frame containing at least the column
named in egoIDcol
, whose values must all be unique. Other columns
contain information about the egos.
For the data.frame
method, it may also contain the
information about the alters in a ‘wide’ format, in the form
of additional columns with names like ATTRNAME1
,
ATTRNAME2
, etc. for attribute ATTRNAME
of alter 1, 2,
etc., as well as a column containing the number of alters nominated
by that ego.
Column name to use for the within-ego index of the alter.
Column name to use for the unique ID of each alter, constructed by concatenating the ID of the ego that nominated them and their index within that ego.
An egodata
object. The object is a list containing the
following elements:
A data frame with one row for each ego, containing at least the
column named in egoIDcol
, and other columns containing attributes of
the egos.
A data frame containing at least the column named in
egoIDcol
, and other columns containing attributes of the alters.
A vector of the same length as the number of egos, containing the relative sampling weight of each ego.
Name of the column in the ego table containing the unique ego identifier.
The following “standard” methods have
also been implemented for egodata
:
A vector with three elements containing the
“dimensions” of the egodata
object: number of egos,
number of columns in the egos
table, and number of columns in the
alters
table, inclsive of the ego identifier column. As a corollary,
nrow
returns the number of egos in the dataset.
A list with three elements containing
the “dimension names” of the egodata
object: ego IDs,
column names of the egos
table, and column names of the alters
table, inclsive of the ego identifier column.
As sample
, but takes and
returns a simulated egodata
dataset by resampling egos,
adjusting ego weights as necessary, if weighted sampling was used.
As head
, but returns the
first n
rows of egos, alters, and weights.
As na.omit.data.frame
,
but takes and returns an egodata
dataset, with egos with
NA
in their rows or in their alters' rows. An optional argument
relevant
, defaulting to all columns, can be used to select (by index
or name) based on which columns an ego may be dropped. (I.e., NA
s in
those not “relevant” are ignored.)
ergm.ego
for examples,
as.network.egodata
, as.egodata.network
,
subset.egodata
, [.egodata