regtable_stack: Stack Multiple Regression Outputs
In yixinsun1216/catwalk: Showing Off Models

This routine takes in multiple outputs from regtable and arranges them into one output. The function stacks the different coefficients and summary statistics, while the extra_rows are grouped at the bottom of the regtable_stack() output.

1 2	regtable_stack(final_tables, table_names = NULL, output_format = "latex", note = NULL, header = NULL)

`final_tables`	list of output produced by regtable, which includes output: a dataframe version of the regtable results model_names: either the character vector specified by mnames in regtable or a function generated vector of numeric IDs for the models To output a dataframe in regtable, pass in "df" for the output_format.
`table_names`	A character vector specifying the names of the objects in final_tables. These names are concatenated with the coefficent names in the output.
`output_format`	A string passed to kable() that specifies the format of the table output. The options are latex, html, markdown, pandoc, and rst. The default is latex.
`note`	A character string if a footnote is to be added to the end of the table.
`header`	A character vector to be passed into `add_header_above` that creates a new header row. This should have length equal to ms from regtable()

# create covariates
x1 <- rnorm(1000)
x2 <- rnorm(length(x1))

# fixed effects
fe <- factor(sample(20, length(x1), replace=TRUE))

# effects for fe
fe_effs <- rnorm(nlevels(fe))

# creating left hand sides y1 and y2
u <- rnorm(length(x1))
y1 <- 2 * x1 + x2 + fe_effs[fe] + u
y2 <- 3 * x1 + x2 + fe_effs[fe] + u

m1 <- felm(y1 ~ x1 + x2 | fe)
m2 <- lm(y1 ~ x1 + x2)

n1 <- felm(y2 ~ x1 + x2 | fe)
n2 <- lm(y2 ~ x1 + x2)

# generate output from regtable
r1 <- regtable(list(m1, m2), est = "x1", 
       output_format = "df")
r2 <- regtable(list(n1, n2), est = "x1", 
       output_format = "df")

regtable_stack(list(r1, r2), table_names = c("1", "2"), output_format = "rst")