如何在texreg调用中自定义拟合优度（gof）？
我有一些要显示的plm模型，但是我只能得到"Num. obs.", "Adj. R^2",和, "R^2"（请参见下面的工作示例）。我想全部显示小n，T，F-statistic和p-value，这些都是我在默认summary()调用中得到的。
我得到的一个例子。首先是一些数据和所需的软件包，
# install.packages(c("wooldridge", "plm", "texreg"), dependencies = TRUE)
library(wooldridge)
data(wagepan)
library(plm)
其次，一些型号
POLS <- plm(lwage ~ educ + black + hisp + exper+I(exper^2)+ married + union +
factor(year), data = wagepan, index=c("nr","year") , model="pooling")
RE <- plm(lwage ~ educ + black + hisp + exper + I(exper^2) + married + union +
factor(year), data = wagepan, index = c("nr","year") , model = "random")
FE <- plm(lwage ~ I(exper^2) + married + union + factor(year),
data = wagepan, index = c("nr","year"), model="within")
第三，我当前的texreg调用及其输出，
# library(texreg)
texreg::screenreg(list(POLS, RE, FE), custom.coef.map = list('married' = 'Marrtied', 'union' = 'Union'))
#> ================================================
#> Model 1 Model 2 Model 3
#> ------------------------------------------------
#> Marrtied 0.11 *** 0.06 *** 0.05 *
#> (0.02) (0.02) (0.02)
#> Union 0.18 *** 0.11 *** 0.08 ***
#> (0.02) (0.02) (0.02)
#> ------------------------------------------------
#> R^2 0.19 0.18 0.18
#> Adj. R^2 0.19 0.18 0.06
#> Num. obs. 4360 4360 4360
#> ================================================
#> *** p < 0.001, ** p < 0.01, * p < 0.05
我确实尝试添加
, include.fstatistic = TRUE，但是似乎无法以这种方式获取它。因为我需要一些其他的自定义。我的目标是这样的，
#> ------------------------------------------------
#> Obs. (N) 4360 4360 4360
#> Indiv.(n) 545 545 545
#> Time (T) 8 8 8
#> R^2 0.19 0.18 0.18
#> Adj. R^2 0.19 0.18 0.06
#> F-stat 72.458 68.4124 83.8515
#> P-value (2.22e-16) (2.22e-16) (2.22e-16)
#> ================================================
#> *** p < 0.001, ** p < 0.01, * p < 0.05
最佳答案
您可以使用texreg::extract()对其进行破解。
为了获得“ small n”，我们首先需要一个小的函数。
getIndex <- function(fit){
# extracts number of factor levels of index variables
# from raw data used in models
index.names <- as.character(as.list(summary(fit)$call)$index)[-1]
if (length(index.names == 1)){
df.name <- as.character(as.list(summary(fit)$call)$data)
index.df <- get(df.name)[, index.names]
length(unique(index.df))
}
if (length(index.names == 2)){
df.name <- as.character(as.list(summary(fit)$call)$data)
index.df <- get(df.name)[, index.names]
cbind(length(unique(index.df[, 1])),
length(unique(index.df[, 2])))
} else {
stop("no index variables specified in model")
}
}
然后继续提取。
fv.1 <- summary(POLS)$fstatistic$statistic # get F statistic
pv.1 <- summary(POLS)$fstatistic$p.value # get p value
ns.1 <- getIndex(POLS)[1] # get small n
tm.1 <- getIndex(POLS)[2] # get times
library(texreg)
ex.1 <- extract(POLS) # extract coefficients and GOF measures
[email protected] <- c([email protected][1:3],"Indiv.(n)", "Time (T)",
"F-stat", "P-value") # the GOF names
ex.1@gof <- c(ex.1@gof[1:3], ns.1, tm.1, fv.1, pv.1) # the GOF values
[email protected] <- c([email protected][1:3], FALSE, FALSE, TRUE, TRUE) # numeric or integer
fv.2 <- summary(RE)$fstatistic$statistic
pv.2 <- summary(RE)$fstatistic$p.value
ns.2 <- getIndex(RE)[1]
tm.2 <- getIndex(RE)[2]
ex.2 <- extract(RE)
[email protected] <- c([email protected][1:3],"Indiv.(n)", "Time (T)",
"F-stat", "P-value")
ex.2@gof <- c(ex.2@gof[1:3], ns.2, tm.2, fv.2, pv.2)
[email protected] <- c([email protected][1:3], FALSE, FALSE, TRUE, TRUE)
fv.3 <- summary(FE)$fstatistic$statistic
pv.3 <- summary(FE)$fstatistic$p.value
ns.3 <- getIndex(FE)[1]
tm.3 <- getIndex(FE)[2]
ex.3 <- extract(FE)
[email protected] <- c([email protected][1:3],"Indiv.(n)", "Time (T)",
"F-stat", "P-value")
ex.3@gof <- c(ex.3@gof[1:3], ns.3, tm.3, fv.3, pv.3)
[email protected] <- c([email protected][1:3], FALSE, FALSE, TRUE, TRUE)
屈服
> screenreg(list(ex.1, ex.2, ex.3))
=======================================================
Model 1 Model 2 Model 3
-------------------------------------------------------
[TRUNCATED...]
-------------------------------------------------------
R^2 0.19 0.18 0.18
Adj. R^2 0.19 0.18 0.06
Num. obs. 4360 4360 4360
Indiv.(n) 545 545 545
Time (T) 8 8 8
F-stat 72.46 68.41 83.85
P-value 0.00 0.00 0.00
=======================================================
*** p < 0.001, ** p < 0.01, * p < 0.05
看一下
str(extract(FE))将此应用于其他GOF。要将其包装为函数，请查看@Neal Fultz的答案中的代码。