In s3alfisc/fwildclusterboot: Fast Wild Cluster Bootstrap Inference for Linear Models

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(fwildclusterboot)
library(fixest)
set.seed(34345)
dqrng::dqset.seed(123)

data(voters)

boottest() supports multiple features of the advanced formula syntax of the fixest package, as e.g. multiple estimations.

Variables created via 'i()'

feols_fit <- feols(proposition_vote ~ i(treatment, ideology1) ,
    data = voters
)

boot1 <- boottest(feols_fit,
    B = 9999,
    param = "treatment::0:ideology1",
    clustid = "group_id1"
)

boot1

Multiple Estimations

It is possible to loop through different regression specifications. For example, to run boottest() over a fixest_multi object created via the sw() function:

feols_fits <- fixest::feols(
  proposition_vote ~ treatment | sw(Q1_immigration, Q2_defense),
  data = voters
)

boot2 <- lapply(
  feols_fits,
  \(x) boottest(
    x,
    B = 999,
    param = "treatment",
    clustid = "group_id1")
) 

boot2

Similarly, it is possible to loop over objects of fixest_multi created by subset functionality:

voters$split <- sample(1:2, nrow(voters), TRUE)

feols_fits <- fixest::feols(
  proposition_vote ~ treatment, 
  split = ~split, 
  data = voters
)

boot3 <- lapply(
  feols_fits, 
  \(x) boottest(
    x,
    B = 999, 
    param = "treatment",
    clustid = "group_id1"
    )
)  

boot3

And of course it is also possible to combine multiple "syntactic sugar" features of fixest:

feols_fits <- fixest::feols(
  proposition_vote ~ treatment | sw(Q1_immigration, Q2_defense), 
  split = ~split,
  data = voters
)

boot4 <- lapply(
  feols_fits, 
  \(x) boottest(
    x, 
    B = 999,
    param = "treatment", 
    clustid = "group_id1"
    )
)  

boot4

What currently is not supported

As a rule of thumb, every transformation of fixed effects is not supported. This includes interaction via ^ and varying slopes syntax via var1[var2]. If you try this, you should receive an error message, and if not, please open an issue on github =) .

Fixed effect interactions via ^ lead to an error. Fixing this is on my backlog but currently not highest priortiy. Send me a message if this causes you large headaches!

feols_fit2 <- feols(proposition_vote ~ treatment | Q1_immigration^Q2_defense,
    data = voters
)

# leads to an error
boot1 <- boottest(feols_fit2,
    B = 9999,
    param = "treatment",
    clustid = "group_id1"
)
# Error in `check_boottest_args_plus()` at fwildclusterboot/R/boottest_fixest.R:514:2:
# ! Advanced formula notation in fixest / fixest via ^ to interact
# fixed effects is currently not supported in boottest().

Relatedly, varying slopes fixed effects syntax is not supported:

feols_fit3 <- feols(proposition_vote ~ treatment | Q1_immigration[log_income],
    data = voters
)

# leads to an error
boot <- boottest(feols_fit3,
    B = 9999,
    param = "treatment",
    clustid = "group_id1"
)
# Error in `check_boottest_args_plus()` at fwildclusterboot/R/boottest_fixest.R:514:2:
# ! Varying slopes in fixest / fixest via [] to interact
# fixed effects is currently not supported in boottest().

s3alfisc/fwildclusterboot documentation built on Sept. 17, 2023, 5:55 a.m.