Generates a self-contained script that loads the data, reads the
specifications, fits them with run_specs() and prints the results
table. By default the specifications are embedded in the script as YAML,
so the one file reproduces the results; pass spec_file to refer to a
separate YAML file instead (see write_specs_yaml()).
Arguments
- specs
A list of
sdid_spec()s (or a single one).- path
File to write, or
NULLto return the script as text.- data_code
R code, as a string, that creates an
sdid_panel()calledpanel. For example'panel <- sdid_panel(read.csv("my_data.csv"), unit = "state", time = "year", outcome = "y")'.- title
A title for the script's header comment.
- spec_file
Path the script should read the specifications from, instead of embedding them.
Details
Each specification is fitted with its own seed (seed_for()), so the
script reproduces every standard error regardless of order.
Examples
code <- write_spec_script(
list(sdid_spec(label = "Prop 99", inference = list(replications = 50))),
path = NULL,
data_code = "panel <- prop99_panel()",
title = "Proposition 99"
)
cat(code)
#> # Proposition 99
#> # Generated by sdidlab 0.1.0 on 2026-09-29.
#> # Each specification is fitted with its own seed (seed_for()), so every
#> # standard error reproduces exactly, whatever order the specs are run in.
#>
#> library(sdidlab)
#>
#> # -- Data ---------------------------------------------------------------
#> panel <- prop99_panel()
#>
#> # -- Specifications -----------------------------------------------------
#> spec_yaml <- r"---(
#> # Proposition 99
#> # Placebo standard errors depend on the RNG stream: fit these in file
#> # order under one seed, or with run_specs(seed_each = TRUE).
#> defaults:
#> inference:
#> replications: 50
#> specs:
#> - label: Prop 99
#> )---"
#> spec_path <- tempfile(fileext = ".yml")
#> writeLines(spec_yaml, spec_path)
#> specs <- read_specs(spec_path)
#>
#> # -- Fit ------------------------------------------------------------------
#> runs <- run_specs(specs, panel, seed_each = TRUE)
#> results <- sdid_results_table(runs)
#> print(results, row.names = FALSE)