Set of functions to directly extract some commonly used statistics, like the p-value or the table of coefficients, from estimations. This was first implemented for fixest estimations, but has some support for other models.

coeftable(
  object,
  vcov = NULL,
  ssc = NULL,
  cluster = NULL,
  keep,
  drop,
  order,
  ...
)

pvalue(object, vcov = NULL, ssc = NULL, cluster = NULL, keep, drop, order, ...)

tstat(object, vcov = NULL, ssc = NULL, cluster = NULL, keep, drop, order, ...)

se(object, vcov = NULL, ssc = NULL, cluster = NULL, keep, drop, order, ...)

Arguments

object

An estimation. For example obtained from feols.

vcov

Versatile argument to specify the VCOV. In general, it is either a character scalar equal to a VCOV type, either a formula of the form: vcov_type ~ variables. The VCOV types implemented are: "iid", "hetero" (or "HC1"), "cluster", "twoway", "NW" (or "newey_west"), "DK" (or "driscoll_kraay"), and "conley". It also accepts object from vcov_cluster, vcov_NW, NW, vcov_DK, DK, vcov_conley and conley. It also accepts covariance matrices computed externally. Finally it accepts functions to compute the covariances. See the `vcov` documentation in the vignette.

ssc

An object of class ssc.type obtained with the function ssc. Represents how the degree of freedom correction should be done.You must use the function ssc for this argument. The arguments and defaults of the function ssc are: adj = TRUE, fixef.K="nested", cluster.adj = TRUE, cluster.df = "min", t.df = "min", fixef.force_exact=FALSE). See the help of the function ssc for details.

cluster

[Fixest specific.] Tells how to cluster the standard-errors (if clustering is requested). Can be either a list of vectors, a character vector of variable names, a formula or an integer vector. Assume we want to perform 2-way clustering over var1 and var2 contained in the data.frame base used for the estimation. All the following cluster arguments are valid and do the same thing: cluster = base[, c("var1, "var2")]}, \code{cluster = c("var1, "var2"), cluster = ~var1+var2. If the two variables were used as clusters in the estimation, you could further use cluster = 1:2 or leave it blank with se = "twoway" (assuming var1 [resp. var2] was the 1st [res. 2nd] cluster).

keep

Character vector. This element is used to display only a subset of variables. This should be a vector of regular expressions (see regex help for more info). Each variable satisfying any of the regular expressions will be kept. This argument is applied post aliasing (see argument dict). Example: you have the variable x1 to x55 and want to display only x1 to x9, then you could use keep = "x[[:digit:]]$". If the first character is an exclamation mark, the effect is reversed (e.g. keep = "!Intercept" means: every variable that does not contain “Intercept” is kept). See details.

drop

Character vector. This element is used if some variables are not to be displayed. This should be a vector of regular expressions (see regex help for more info). Each variable satisfying any of the regular expressions will be discarded. This argument is applied post aliasing (see argument dict). Example: you have the variable x1 to x55 and want to display only x1 to x9, then you could use drop = "x[[:digit:]]{2}". If the first character is an exclamation mark, the effect is reversed (e.g. drop = "!Intercept" means: every variable that does not contain “Intercept” is dropped). See details.

order

Character vector. This element is used if the user wants the variables to be ordered in a certain way. This should be a vector of regular expressions (see regex help for more info). The variables satisfying the first regular expression will be placed first, then the order follows the sequence of regular expressions. This argument is applied post aliasing (see argument dict). Example: you have the following variables: month1 to month6, then x1 to x5, then year1 to year6. If you want to display first the x's, then the years, then the months you could use: order = c("x", "year"). If the first character is an exclamation mark, the effect is reversed (e.g. order = "!Intercept" means: every variable that does not contain “Intercept” goes first). See details.

...

Other arguments to be passed to summary.

se

[Fixest specific.] Character scalar. Which kind of standard error should be computed: “iid”, “hetero”, “cluster”, “twoway”, “threeway” or “fourway”? By default if there are fixed-effects in the estimation: se = "cluster", otherwise se = "iid". Note that this argument is not needed if the argument cluster is present.

Value

Returns a table of coefficients, with in rows the variables and four columns: the estimate, the standard-error, the t-statistic and the p-value.

Details

This set of functions is primarily constructed for fixest estimations. Although it can work for non-fixest estimations, support for exotic estimation procedures that do not report standardized coefficient tables is highly limited.

Functions

  • pvalue: Extracts the p-value of an estimation

  • tstat: Extracts the t-statistics of an estimation

  • se: Extracts the standard-error of an estimation

Examples


# Some data and estimation
data(trade)
est = fepois(Euros ~ log(dist_km) | Origin^Product + Year, trade)

#
# Coeftable/se/tstat/pvalue
#

# Default is clustering along Origin^Product
coeftable(est)
#>               Estimate Std. Error   t value      Pr(>|t|)
#> log(dist_km) -1.023957 0.04728994 -21.65275 5.725404e-104
#> attr(,"type")
#> [1] "Clustered (Origin^Product)"
se(est)
#> log(dist_km) 
#>   0.04728994 
tstat(est)
#> log(dist_km) 
#>    -21.65275 
pvalue(est)
#>  log(dist_km) 
#> 5.725404e-104 

# Now with two-way clustered standard-errors
#  and using coeftable()

coeftable(est, cluster = ~Origin + Product)
#>               Estimate Std. Error   t value    Pr(>|t|)
#> log(dist_km) -1.023957  0.0906375 -11.29728 1.35342e-29
#> attr(,"type")
#> [1] "Clustered (Origin & Product)"
se(est, cluster = ~Origin + Product)
#> log(dist_km) 
#>    0.0906375 
pvalue(est, cluster = ~Origin + Product)
#> log(dist_km) 
#>  1.35342e-29 
tstat(est, cluster = ~Origin + Product)
#> log(dist_km) 
#>    -11.29728 

# Or you can cluster only once:
est_sum = summary(est, cluster = ~Origin + Product)
coeftable(est_sum)
#>               Estimate Std. Error   t value    Pr(>|t|)
#> log(dist_km) -1.023957  0.0906375 -11.29728 1.35342e-29
#> attr(,"type")
#> [1] "Clustered (Origin & Product)"
se(est_sum)
#> log(dist_km) 
#>    0.0906375 
tstat(est_sum)
#> log(dist_km) 
#>    -11.29728 
pvalue(est_sum)
#> log(dist_km) 
#>  1.35342e-29 

# You can use the arguments keep, drop, order
# to rearrange the results

base = iris
names(base) = c("y", "x1", "x2", "x3", "species")

est_iv = feols(y ~ x1 | x2 ~ x3, base)

tstat(est_iv, keep = "x1")
#>       x1 
#> 7.960735 
coeftable(est_iv, keep = "x1|Int")
#>             Estimate Std. Error  t value     Pr(>|t|)
#> (Intercept) 2.438955 0.25349903 9.621160 2.688392e-17
#> x1          0.559183 0.07024264 7.960735 4.261663e-13

coeftable(est_iv, order = "!Int")
#>              Estimate Std. Error   t value     Pr(>|t|)
#> fit_x2      0.4509765 0.01794806 25.126759 4.556383e-55
#> x1          0.5591830 0.07024264  7.960735 4.261663e-13
#> (Intercept) 2.4389548 0.25349903  9.621160 2.688392e-17