R-devel/estout_rdrobust.R

In jamesfeigenbaum/textablr: Output Regression Results and Other Tables to LaTeX

# library(tidyverse)
# library(broom)
# library(stringr)
# library(magrittr)
# library(purrr)
# library(data.table)
# library(jjfPkg)
#
# # compare with stargazer
# library(stargazer)
#
# # regression packages
# library(lfe)
# library(rdrobust)
#
# rm(list = ls())
#
# x <- runif(1000, -5, 5)
# y <- 5 + 2*x + 3*(x >= 0) + rnorm(1000)
# dt <- tibble(x = x, y = y)
#
# ggplot(dt, aes(x = x, y = y, group = (x>0))) +
#   geom_point(alpha = .1) +
#   geom_smooth(se = FALSE) +
#   geom_vline(xintercept = 0, color = "red", linetype = "dashed") +
#   theme_jjf()
#
# # create dummy data for RD
#
# reg1 <- rdrobust(dt$y, dt$x, )
# reg2 <- rdrobust(dt$y, dt$x, p = 2)
# reg3 <- rdrobust(dt$y, dt$x, kernel = "epanechnikov")
# reg4 <- rdrobust(dt$y, dt$x, kernel = "uniform")
#
# # input
#
# # function will take regressions as a list of list objects
# regs <- list(reg1, reg2, reg3, reg4)
#
# # one optional parameter will be the name of the running variable
# var_labels <- "Running Variable X"
#
# # which summary stats to include?
# sumstat_include <- c("N", "h_l", "h_r", "b_l", "b_r", "N_h_l", "N_h_r", "c", "p", "q",
#                      "kernel", "vce", "bwselect")
#
# source("R/lookups.R")
#
# # we're going to extract some things right from regs
# # but other things from the list of summaries
# # regs_summary <- map(regs, summary)
# # doesn't work for rdrobust
#
# # simple stuff
# # all with the prefix regs_
#
# # count the number of regressions
# # for numbering columns
# reg_columns <- length(regs)
#
# # get the functions called to run regressions
# # useful metadata
# # TODO add this to output
# reg_calls <- map(regs, extract2, c("call")) %>%
#   as.character()
#
# # summary statistics
# source("R/sumstat_rdrobust.R")
# out_sumstats <- sumstat_master(regs, sumstat_include, sumstat_names_rdrobust)
#
# source("R/coeffs.R")
# out_x_fe <- x_fe_master(regs, var_labels, var_indicates, var_omits)
#
# out_colnumbers <- 1:reg_columns %>%
#   # surround with \multicolumn{1}{c}{XXX}
#   sprintf("\\multicolumn{1}{c}{(%s)}", .) %>%
#   map(extract) %>%
#   purrr::transpose() %>%
#   map_chr(paste0, collapse = " & ") %>%
#   str_c("& ", ., " \\\\")
#
# out_table <- c(
#   out_colnumbers,
#   "\\midrule",
#   out_x_fe,
#   "\\midrule",
#   out_sumstats)
#
# # TODO
# # if the following row has a midrule, remove the previous addlinespace
#
# cat(out_table, file = "latex_testing/table1.tex", sep = "\n")