Debian Codegraph @

man/vcovCR.lm_robust.Rd

raw
1 % Generated by roxygen2: do not edit by hand
2 % Please edit documentation in R/lm_robust.R
3 \name{vcovCR.lm_robust}
4 \alias{vcovCR.lm_robust}
5 \title{Cluster-robust variance-covariance matrix for an \code{estimatr::lm_robust}
6 object.}
7 \usage{
8 \method{vcovCR}{lm_robust}(
9 obj,
10 cluster,
11 type,
12 target = NULL,
13 inverse_var = NULL,
14 form = "sandwich",
15 ...
16 )
17 }
18 \arguments{
19 \item{obj}{Fitted model for which to calculate the variance-covariance matrix}
20
21 \item{cluster}{Expression or vector indicating which observations belong to
22 the same cluster. If not specified, will be detected from the
23 \code{clusters} argument of \code{obj}.}
24
25 \item{type}{Character string specifying which small-sample adjustment should
26 be used, with available options \code{"CR0"}, \code{"CR1"}, \code{"CR1p"},
27 \code{"CR1S"}, \code{"CR2"}, or \code{"CR3"}. If not specified, will be
28 detected from the \code{se_type} argument of \code{obj}. See "Details"
29 section of \code{\link{vcovCR}} for further information.}
30
31 \item{target}{Optional matrix or vector describing the working
32 variance-covariance model used to calculate the \code{CR2} and \code{CR4}
33 adjustment matrices. If a vector, the target matrix is assumed to be
34 diagonal. If not specified, the target is taken to be an identity matrix.}
35
36 \item{inverse_var}{Optional logical indicating whether the weights used in
37 fitting the model are inverse-variance. If not specified, \code{vcovCR}
38 will attempt to infer a value.}
39
40 \item{form}{Controls the form of the returned matrix. The default
41 \code{"sandwich"} will return the sandwich variance-covariance matrix.
42 Alternately, setting \code{form = "meat"} will return only the meat of the
43 sandwich and setting \code{form = B}, where \code{B} is a matrix of
44 appropriate dimension, will return the sandwich variance-covariance matrix
45 calculated using \code{B} as the bread. \code{form = "estfun"} will return the
46 (appropriately scaled) estimating function, the transposed crossproduct of
47 which is equal to the sandwich variance-covariance matrix.}
48
49 \item{...}{Additional arguments available for some classes of objects.}
50 }
51 \value{
52 An object of class \code{c("vcovCR","clubSandwich")}, which consists
53 of a matrix of the estimated variance of and covariances between the
54 regression coefficient estimates.
55 }
56 \description{
57 \code{vcovCR} returns a sandwich estimate of the variance-covariance matrix
58 of a set of regression coefficient estimates from an
59 \code{\link[estimatr]{lm_robust}} object.
60 }
61 \examples{
62
63 data("ChickWeight", package = "datasets")
64 ChickWeight$Chick <- factor(ChickWeight$Chick, ordered = FALSE)
65
66 if (requireNamespace("estimatr", quietly = TRUE)) withAutoprint({
67 library(estimatr)
68
69 lm_fit <- lm_robust(
70 weight ~ Time + Diet:Time,
71 data = ChickWeight
72 )
73 vcovCR(lm_fit, cluster = ChickWeight$Chick, type = "CR2")
74
75 lm_fit_clust <- lm_robust(
76 weight ~ Time + Diet:Time, data = ChickWeight,
77 clusters = Chick
78 )
79 conf_int(lm_fit_clust, vcov = "CR2")
80
81 # similar model via lm_lin()
82 lin_fit_clust <- lm_lin(
83 weight ~ Diet,
84 covariates = ~ Time,
85 data = ChickWeight,
86 clusters = Chick
87 )
88 conf_int(lin_fit_clust, vcov = "CR2")
89
90 lm_fit_fe <- lm_robust(
91 weight ~ Time:Diet, data = ChickWeight,
92 clusters = Chick,
93 fixed_effects = ~ Chick
94 )
95 vcovCR(lm_fit_fe)
96
97 # two-way fixed effects model
98 data("MortalityRates")
99 MortalityRates <- subset(MortalityRates, cause == "Motor Vehicle")
100 MortalityRates$state <- factor(MortalityRates$state)
101 MortalityRates$year <- factor(MortalityRates$year)
102 MLDA_fit <- lm_robust(
103 mrate ~ legal + beertaxa + beerpercap + winepercap + spiritpercap,
104 fixed_effects = ~ year + state,
105 data = MortalityRates,
106 cluster = state
107 )
108 conf_int(MLDA_fit, vcov = "CR2")
109
110 if (requireNamespace("plm", quietly = TRUE)) withAutoprint({
111
112 data("Produc", package = "plm")
113 lm_individual <- lm_robust(
114 log(gsp) ~ log(pcap) + log(pc) + log(emp) + unemp,
115 data = Produc,
116 fixed_effects = ~ state,
117 cluster = state
118 )
119 vcovCR(lm_individual, type = "CR2")
120
121 })
122
123 })
124
125 }
126 \seealso{
127 \code{\link{vcovCR}}
128 }