Help for package tidysynthesis

Title:

A Common API for Synthesizing Data

Version:

0.1.2

Maintainer:

Aaron R. Williams <awilliams@urban.org>

Description:

A system built on 'tidymodels' for generating synthetic tabular data. We provide tools for ordering a sequential synthesis, feature and target engineering, sampling, hyperparameter tuning, enforcing constraints, and adding extra noise during a synthesis.

URL:

https://ui-research.github.io/tidysynthesis-documentation/

BugReports:

https://github.com/UrbanInstitute/tidysynthesis/issues

Depends:

R (≥ 4.1.0)

Imports:

dplyr, forcats, parsnip, pillar, purrr, progressr, recipes, rlang, rsample, stringr, tibble, tidyr (≥ 1.0.0), tune, vctrs, workflows, yardstick, ExtDist, dapper

Suggests:

hardhat, palmerpenguins, poissonreg, randomForest, ranger, rpart, rpart.LAD (≥ 0.1.2), testthat (≥ 2.1.0), usethis, VGAM

License:

GPL-3

Encoding:

UTF-8

LazyData:

true

RoxygenNote:

7.3.2

NeedsCompilation:

Packaged:

2025-11-06 15:30:08 UTC; aaronwilliams

Author:

Aaron R. Williams

[aut, cre], Kyle Ueyama [aut], Gabe Morrison

[aut], Jeremy Seeman

[aut], Elyse McFalls [ctb], Claire Morton [ctb], Livia Mucciolo [ctb], Madeline Pickens [ctb], Noah Zwiefel [ctb], The Urban Institute [cph]

Repository:

CRAN

Date/Publication:

2025-11-11 21:20:07 UTC

constant mapping between component names and inspections

Description

constant mapping between component names and inspections

Usage

.name_to_inspect

Format

An object of class list of length 7.

American Community Survey confidential microdata (with weights)

Description

An extract constructed from the 2019 American Community Survey containing a survey sample of n = 1500 Nebraska respondents, with survey weights included.

Usage

acs_conf

Format

`acs_conf`

A data frame with 1,500 rows and 12 columns:

county: fct, county
gq: fct, group quarter kind
sex: fct, sex
marst: fct, marital status
hcovany: fct, health insurance status
empstat: fct, employment status; contains empty levels.
classwkr: fct, employment kind (ex: self-employed, etc.); contains "N/A" levels.
age: dbl, age (in years)
famsize: dbl, household/family size
transit_time: dbl, transit time to work (in minutes)
inctot: dbl, annual income; contains missing values
wgt: dbl, survey weight

Details

Original data source: Steven Ruggles, Sarah Flood, Matthew Sobek, Daniel Backman, Annie Chen, Grace Cooper, Stephanie Richards, Renae Rogers, and Megan Schouweiler. IPUMS USA: Version 15.0 [dataset]. Minneapolis, MN: IPUMS, 2024. https://doi.org/10.18128/D010.V15.0

Source

https://usa.ipums.org/usa/

American Community Survey confidential microdata (without weights)

Description

An extract constructed from the 2019 American Community Survey containing a survey sample of n = 1500 Nebraska respondents, with survey weights included.

Usage

acs_conf_nw

Format

`acs_conf_nw`

A data frame with 1,500 rows and 11 columns:

county: fct, county
gq: fct, group quarter kind
sex: fct, sex
marst: fct, marital status
hcovany: fct, health insurance status
empstat: fct, employment status; contains empty levels.
classwkr: fct, employment kind (ex: self-employed, etc.); contains "N/A" levels.
age: dbl, age (in years)
famsize: dbl, household/family size
transit_time: dbl, transit time to work (in minutes)
inctot: dbl, annual income; contains missing values

Details

Source

https://usa.ipums.org/usa/

American Community Survey starting microdata (with weights)

Description

An extract constructed from the 2019 American Community Survey containing a survey sample of n = 500 Nebraska respondents, with survey weights included.

Usage

acs_start

Format

`acs_start`

A data frame with 500 rows and 5 columns:

county: fct, county
gq: fct, group quarter kind
sex: fct, sex
marst: fct, marital status
wgt: dbl, survey weight

Details

Source

https://usa.ipums.org/usa/

American Community Survey starting microdata (without weights)

Description

An extract constructed from the 2019 American Community Survey containing a survey sample of n = 500 Nebraska respondents, with survey weights included.

Usage

acs_start_nw

Format

`acs_start_nw`

A data frame with 500 rows and 4 columns:

county: fct, county
gq: fct, group quarter kind
sex: fct, sex
marst: fct, marital status

Details

Source

https://usa.ipums.org/usa/

Inject noise into a categorical random variable by mixing a sample of uniform records into the predictions.

Description

Inject noise into a categorical random variable by mixing a sample of uniform records into the predictions.

Usage

add_noise_cat_unif(
  model,
  new_data,
  conf_model_data,
  outcome_var,
  col_schema,
  pred,
  unif_prop,
  resample_props = NULL,
  observed_levels = FALSE
)

Arguments

model

A model_spec or a list of model_specs from library(parsnip)

new_data

A data frame used to generate predictions

conf_model_data

A data frame for estimating the predictive model

outcome_var

A string name representing the outcome variable

col_schema

A list of column schema specifications for the new variable

pred

A vector of values predicted by the model

unif_prop

A proportion of records to resample with uniform noise

resample_props

An optional named vector of probabilities for resampling, defaults to uniform over all levels supplied in col_schema.

observed_levels

An optional Boolean to only resample from observed levels in the confidential data.

Value

A numeric vector with noise added to each prediction

Examples


conf_model_data <- mtcars|>
  dplyr::mutate(gear = factor(.data[["gear"]]))

col_schema <- list(
  "dtype" = "fct",
  "levels" = c("3", "4", "5"),
  "na_prop" = 0
)

add_noise_cat_unif(
  model = conf_model_data,
  new_data = NULL,
  conf_model_data = NULL,
  outcome_var = "gear",
  col_schema = col_schema,
  pred = factor(c(rep("3", 10), rep("4", 10), rep("5", 10))),
  unif_prop = 0.5
)

Add discrete normal noise with mean 0 to predicted values with constant variance

Description

Add discrete normal noise with mean 0 to predicted values with constant variance

Usage

add_noise_disc_gaussian(
  model,
  new_data,
  conf_model_data,
  outcome_var,
  col_schema,
  pred,
  variance = NULL,
  rho = NULL,
  sensitivity = NULL,
  increment = 1
)

Arguments

model

A model_spec or a list of model_specs from library(parsnip)

new_data

A data frame used to generate predictions

conf_model_data

A data frame for estimating the predictive model

outcome_var

A string name representing the outcome variable

col_schema

A list of column schema specifications for the new variable

pred

A vector of values predicted by the model

variance

float, sampling variance for additive noise

rho

float, alternative privacy loss budget prescribed by the Gaussian mechanism under rho-zero-concentrated differential privacy.

sensitivity

float, alternative sample sensitivity prescribed by the Gaussian mechanism under rho-zero-concentrated differential privacy.

increment

Numeric indicating space between discrete noise samples, defaults to 1. Note that this does not impact the noise sampling variance, as the increment rescales noise distributions specified by sampling variance.

Value

A numeric vector with noise added to each prediction

Examples


add_noise_disc_gaussian(
  model = NULL,
  new_data = NULL,
  conf_model_data = NULL,
  outcome_var = NULL,
  col_schema = NULL,
  pred = 1:100,
  variance = 3
)

Add discrete Laplace noise with mean 0 to predicted values with constant variance

Description

Add discrete Laplace noise with mean 0 to predicted values with constant variance

Usage

add_noise_disc_laplace(
  model,
  new_data,
  conf_model_data,
  outcome_var,
  col_schema,
  pred,
  variance = NULL,
  epsilon = NULL,
  sensitivity = NULL,
  increment = 1
)

Arguments

model

A model_spec or a list of model_specs from library(parsnip)

new_data

A data frame used to generate predictions

conf_model_data

A data frame for estimating the predictive model

outcome_var

A string name representing the outcome variable

col_schema

A list of column schema specifications for the new variable

pred

A vector of values predicted by the model

variance

float, sampling variance for additive noise

epsilon

float, alternative privacy loss budget prescribed by the Laplace mechanism under epsilon differential privacy.

sensitivity

float, alternative sample sensitivity prescribed by the Laplace mechanism under epsilon differential privacy.

increment

Value

A numeric vector with noise added to each prediction

Examples


add_noise_disc_laplace(
  model = NULL,
  new_data = NULL,
  conf_model_data = NULL,
  outcome_var = NULL,
  col_schema = NULL,
  pred = 1:100,
  variance = 3
)

Add normal noise with mean 0 to predicted values with constant variance

Description

Add normal noise with mean 0 to predicted values with constant variance

Usage

add_noise_gaussian(
  model,
  new_data,
  conf_model_data,
  outcome_var,
  col_schema,
  pred,
  variance = NULL,
  rho = NULL,
  sensitivity = NULL
)

Arguments

model

A model_spec or a list of model_specs from library(parsnip)

new_data

A data frame used to generate predictions

conf_model_data

A data frame for estimating the predictive model

outcome_var

A string name representing the outcome variable

col_schema

A list of column schema specifications for the new variable

pred

A vector of values predicted by the model

variance

Sampling variance for additive noise

rho

Alternative privacy loss budget prescribed by the Gaussian mechanism under rho-zero-concentrated differential privacy.

sensitivity

Alternative sample sensitivity prescribed by the Gaussian mechanism under rho-zero-concentrated differential privacy.

Value

A numeric vector with noise added to each prediction

Examples


add_noise_gaussian(
  model = NULL,
  new_data = NULL,
  conf_model_data = NULL,
  outcome_var = NULL,
  col_schema = NULL,
  pred = 1:100,
  variance = 3
)

Add normal noise to predicted values with variances calculated for ntiles using Gaussian kernel density estimators

Description

Add normal noise to predicted values with variances calculated for ntiles using Gaussian kernel density estimators

Usage

add_noise_kde(
  model,
  new_data,
  conf_model_data,
  outcome_var,
  col_schema,
  pred,
  exclusions = NULL,
  n_ntiles = NULL,
  obs_per_ntile = NULL,
  ties_method = "collapse",
  sd_scale = 1
)

Arguments

model

A model_spec or a list of model_specs from library(parsnip)

new_data

A data frame used to generate predictions

conf_model_data

A data frame for estimating the predictive model

outcome_var

A string name representing the outcome variable

col_schema

A list of column schema specifications for the new variable

pred

A vector of values predicted by the model

exclusions

Numeric values that should not receive extra noise

n_ntiles

The number of ntiles

obs_per_ntile

A numeric for the minimum number of observations to be in an ntile. Cannot be used in conjunction with the n_ntiles argument.

ties_method

The ntiles approach to adding noise requires a one-to-one mapping from model-generated values to ntiles in the original data. The methods "collapse", "random", and "exclusions" deal with situations where the ntiles lack unique bounds. "collapse" collapses ntile breaks to preserve the one-to-one relationship; "random" adds a small random perturbation to the derived boundaries; finally, "exclusions" treats ntile tie values as derived exclusions.

sd_scale

float, a positive number to scale the estimated KDE variance. Defaults to 1.0

Value

A numeric vector with noise added to each prediction

Examples


add_noise_kde(
  model = NULL,
  new_data = tibble::tibble(x = 1:100),
  conf_model_data = tibble::tibble(x = 1:100),
  outcome_var = "x",
  col_schema = NULL,
  pred = 1:100,
  n_ntiles = 4
)

Add Laplace noise with mean 0 to predicted values with constant variance

Description

Add Laplace noise with mean 0 to predicted values with constant variance

Usage

add_noise_laplace(
  model,
  new_data,
  conf_model_data,
  outcome_var,
  col_schema,
  pred,
  variance = NULL,
  epsilon = NULL,
  sensitivity = NULL
)

Arguments

model

A model_spec or a list of model_specs from library(parsnip)

new_data

A data frame used to generate predictions

conf_model_data

A data frame for estimating the predictive model

outcome_var

A string name representing the outcome variable

col_schema

A list of column schema specifications for the new variable

pred

A vector of values predicted by the model

variance

Sampling variance for additive noise

epsilon

Alternative privacy loss budget prescribed by the Laplace mechanism under epsilon differential privacy.

sensitivity

Alternative sample sensitivity prescribed by the Laplace mechanism under epsilon differential privacy.

Value

A numeric vector with noise added to each prediction

Examples


add_noise_laplace(
  model = NULL,
  new_data = NULL,
  conf_model_data = NULL,
  outcome_var = NULL,
  col_schema = NULL,
  pred = 1:100,
  variance = 3
)

Add to visit sequence for factor variables

Description

Add to visit sequence for factor variables

Usage

add_sequence_factor(roadmap, ..., method = c("entropy", "gini"))

Arguments

roadmap

A roadmap object

...

⁠<tidy-select>⁠ One or more unquoted expressions separated by commas. Variable names can be used as if they were positions in the data frame, so expressions like x:y can be used to select a range of variables.

method

A quoted name for the method used to sort the visit_sequence. Current methods include "entropy" and "gini".

Value

An updated visit_sequence

Examples


roadmap(
  conf_data = acs_conf_nw, 
  start_data = acs_start_nw
) |>
  add_sequence_factor(dplyr::where(is.factor), method = "gini")

Add to visit sequence using a manual method

Description

Add to visit sequence using a manual method

Usage

add_sequence_manual(roadmap, ...)

Arguments

roadmap

A roadmap object.

...

Value

An updated roadmap object.

Examples


roadmap(
  conf_data = acs_conf_nw, 
  start_data = acs_start_nw
) |>
  add_sequence_manual(
    c("inctot", "hcovany", "empstat",  "classwkr",  "age",   "famsize",  
      "transit_time")
  )

Add to visit sequence for numeric variables

Description

Add to visit sequence for numeric variables

Usage

add_sequence_numeric(
  roadmap,
  ...,
  method = c("correlation", "proportion", "weighted total", "absolute weighted total",
    "weighted absolute total"),
  cor_var = NULL,
  na.rm = FALSE,
  cor_use = "everything"
)

Arguments

roadmap

A roadmap object

...

method

A quoted name for the method used to sort the visit_sequence

cor_var

A numeric variable for the correlation method

na.rm

Boolean that if TRUE, removes NA values from computations

cor_use

A string correlation data method passed to stats::cor if using. If na.rm == TRUE then defaults to complete.obs. See ?stats::cor for more options.

Value

An updated visit_sequence

Examples


roadmap(
  conf_data = acs_conf_nw, 
  start_data = acs_start_nw
) |>
  add_sequence_numeric(
    dplyr::where(is.numeric), 
    method = "correlation", 
    cor_var = "age",
    na.rm = TRUE
  )

Collapse data frames with _NA variables to coerce related variables to include NA

Description

Collapse data frames with _NA variables to coerce related variables to include NA

Usage

collapse_na(data)

Arguments

data

A data frame with columns ending in _NA

Value

A data frame with no _NA columns and NA values

Examples


example_na_expanded <- expand_na(data = example_na)

collapse_na(data = example_na_expanded)

Create a constraints object

Description

Create a constraints object

Usage

constraints(
  schema,
  constraints_df_num = NULL,
  constraints_df_cat = NULL,
  max_z_num = 0,
  max_z_cat = 0
)

Arguments

schema

A schema object

constraints_df_num

A specially formatted data frame with constraints to be imposed during the synthesis process. See examples for formatting.

constraints_df_cat

A specifically formatted data frame with constraints to be imposed during the synthesis process.

max_z_num

Numeric vector(s) for the number of times a value should be resampled before hardbounding if it violates a constraint.

max_z_cat

Numeric vector(s) for the number of times a value should be resampled before hardbounding if it violates a constraint.

Value

A constraints object.

Examples


constraints(
  schema = schema(
    conf_data = mtcars |> dplyr::mutate(vs = factor(vs)),
    start_data = dplyr::select(mtcars, cyl)
  ),
  constraints_df_num = tibble::tribble(
    ~var, ~min, ~max, ~conditions, 
    # ensure all mpg values are greater than 0
    "mpg", 0, Inf, "TRUE",
    # ensure when cyl == 6, mpg is less than 15
    "mpg", -Inf, 15, "cyl == 6",
    # ensure disp is always between 0 and 150
    "disp", 0, 150, "TRUE"
  ),
  constraints_df_cat = tibble::tribble(
    ~var, ~allowed, ~forbidden, ~conditions, 
    # ensure vs != 1 when gear >= 4
    "vs", NA,  1, "gear >= 5",
    # ensure vs == 1 when gear >= 4
    "vs", 0,  NA, "gear == 4"
  )
)

Add, update, or reset a `constraints` object within an existing `roadmap`.

Description

Add, update, or reset a constraints object within an existing roadmap.

Usage

add_constraints(roadmap, constraints)

update_constraints(roadmap, ...)

reset_constraints(roadmap)

Arguments

roadmap

A roadmap object

constraints

A constraints object.

...

Optional named parameters passed to constraints().

Value

A new roadmap object.

A roadmap object with added constraints.

A roadmap object with updated constraints.

A roadmap object with reset constraints.

Examples


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

constraints_df_num <- 
  tibble::tribble(~var, ~min, ~max, ~conditions,
                  "transit_time", 0, 300, "TRUE")

constraints <- constraints(
  schema = rm[["schema"]],
  constraints_df_num = constraints_df_num,
  max_z_num = 0
)

rm |>
  add_constraints(constraints)


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

constraints_df_num <- 
  tibble::tribble(~var, ~min, ~max, ~conditions,
                  "transit_time", 0, 300, "TRUE")

constraints <- constraints(
  schema = rm[["schema"]],
  constraints_df_num = constraints_df_num,
  max_z_num = 0
)

rm |>
  update_constraints(constraints)


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

constraints_df_num <- 
  tibble::tribble(~var, ~min, ~max, ~conditions,
                  "transit_time", 0, 300, "TRUE")

constraints <- constraints(
  schema = rm[["schema"]],
  constraints_df_num = constraints_df_num,
  max_z_num = 0
)

rm <- rm |>
  add_constraints(constraints)

reset_constraints(rm)

Construct a list of extractors for parsnip models

Description

Construct a list of extractors for parsnip models

Usage

construct_extractors(
  roadmap,
  default_extractor = NULL,
  custom_extractors = NULL
)

Arguments

roadmap

A roadmap object

default_extractor

An extractor from library(parsnip)

custom_extractors

A formatted list of extractors

Value

A named list of extractors

Examples


# construct_extractors() can create a sequence of extractors using a fully-default 
# approach, a hybrid approach, or a fully-customized approach. All approaches
# require a roadmap and extractors. 

rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

# Fully-default approach

construct_extractors(
  roadmap = rm, 
  default_extractor = parsnip::extract_fit_engine
)

# Hybrid approach

construct_extractors(
  roadmap = rm, 
  default_extractor = parsnip::extract_fit_engine,
  custom_extractors = list(
    list(vars = "hcovany", extractor = parsnip::extract_parameter_dials)
  )
)

# Fully-customized approach

construct_extractors(
  roadmap = rm, 
  custom_extractors = list(
    list(
      vars = c("hcovany", "empstat", "classwkr"), 
      extractor = parsnip::extract_fit_engine
    ),
    list(
      vars = c("age", "famsize", "transit_time", "inctot"), 
      extractor = parsnip::extract_parameter_dials
    )
  )
)

Construct a list of models for synthesis

Description

Construct a list of models for synthesis

Usage

construct_models(
  roadmap,
  default_regression_model = NULL,
  default_classification_model = NULL,
  custom_models = NULL
)

Arguments

roadmap

A roadmap object

default_regression_model

A parsnip model object used for regression in numeric outcome variables

default_classification_model

A parsnip model object used for classification in categorical outcome variables

custom_models

A formatted list with parsnip model objects explicitly paired with every variable in the visit_sequence

Value

A named list of models

Examples


# construct_models() can create a sequence of models using a fully-default 
# approach, a hybrid approach, or a fully-customized approach. All approaches
# require a roadmap and model objects. 

rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

rpart_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

rpart_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "classification")

lm_mod <- parsnip::linear_reg() |> 
  parsnip::set_engine("lm") |>
  parsnip::set_mode(mode = "regression")

# Fully-default approach

construct_models(
  roadmap = rm, 
  default_regression_model = lm_mod, 
  default_classification_model = rpart_mod_class
)

# Hybrid approach

construct_models(
  roadmap = rm, 
  default_regression_model = lm_mod,
  default_classification_model = rpart_mod_class,
  custom_models = list(
    list(vars = "age", model = lm_mod)
  )
)

# Fully-customized approach

construct_models(
  roadmap = rm, 
  custom_models = list(
    list(vars = c("hcovany", "empstat", "classwkr"), model = rpart_mod_class),
    list(vars = c("age", "famsize", "transit_time", "inctot"), model = rpart_mod_reg)
  )
)

Construct a list of noise objects for synthesis

Description

Construct a list of noise objects for synthesis

Usage

construct_noise(
  roadmap,
  default_regression_noise = NULL,
  default_classification_noise = NULL,
  custom_noise = NULL
)

Arguments

roadmap

A roadmap object

default_regression_noise

A noise function for regression models

default_classification_noise

A noise function for classification models

custom_noise

A formatted list of noise functions

Value

A named list of noise

Examples


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

noise_defaults <- construct_noise(
  roadmap = rm,
  default_regression_noise = noise(),
  default_classification_noise = noise()
)


# construct_noise() can create a sequence of noise objects using a 
# fully-default approach, a hybrid approach, or a fully-customized approach. 
# All approaches require a roadmap and noise objects. 

rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

noise_reg <- noise(
  add_noise = TRUE,
  mode = "regression",
  noise_fun = add_noise_gaussian
)

noise_class <- noise(
  add_noise = TRUE,
  mode = "classification",
  noise_fun = add_noise_cat_unif
)

# Fully-default approach

construct_noise(
  roadmap = rm, 
  default_regression_noise = noise_reg, 
  default_classification_noise = noise_class
)

# Hybrid approach

noise_reg2 <- noise(
  add_noise = TRUE,
  mode = "regression",
  noise_fun = add_noise_disc_gaussian 
)

construct_noise(
  roadmap = rm, 
  default_regression_noise = noise_reg,
  default_classification_noise = noise_class,
  custom_noise = list(
    list(vars = "age", noise = noise_reg2)
  )
)

# Fully-customized approach

construct_noise(
  roadmap = rm, 
  custom_noise = list(
    list(vars = c("hcovany", "empstat", "classwkr"), noise = noise_class),
    list(vars = c("age", "famsize", "transit_time", "inctot"), noise = noise_reg)
  )
)

Construct a sequence of model recipes for sequential synthesis

Description

Construct a sequence of model recipes for sequential synthesis

Usage

construct_recipes(
  roadmap,
  default_regression_steps = NULL,
  default_classification_steps = NULL,
  custom_steps = NULL
)

Arguments

roadmap

A roadmap object

default_regression_steps

A list containing one or more recipes::step_*()

default_classification_steps

A list containing one or more recipes::step_*()

custom_steps

A list of lists containing one or more recipes::step_*()

Value

A list of formulas

Examples


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

construct_recipes(rm)


# construct_recipes() can create a sequence of recipes using a fully-default 
# approach, a hybrid approach, or a fully-customized approach. All approaches
# require a roadmap and steps. 

rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

step1 <- function(x) {
x |>
  recipes::step_center(recipes::all_predictors(), id = "center")
}

# Fully-default approach

construct_recipes(
  roadmap = rm, 
  default_regression_steps = step1, 
  default_classification_steps = step1
)

# Hybrid approach

step2 <- function(x) {
  x |>
    recipes::step_scale(recipes::all_predictors(), id = "scale")
}

construct_recipes(
  roadmap = rm, 
  default_regression_steps = step1,
  default_classification_steps = step1,
  custom_steps = list(
    list(vars = "age", step = step2)
  )
)

# Fully-customized approach

construct_recipes(
  roadmap = rm, 
  custom_steps = list(
    list(vars = c("hcovany", "empstat", "classwkr"), step = step1),
    list(vars = c("age", "famsize", "transit_time", "inctot"), step = step1)
  )
)

Construct a list of samplers for synthesis

Description

Construct a list of samplers for synthesis

Usage

construct_samplers(
  roadmap,
  default_regression_sampler = NULL,
  default_classification_sampler = NULL,
  custom_samplers = NULL
)

Arguments

roadmap

A roadmap object

default_regression_sampler

A sampler function for regression models

default_classification_sampler

A sampler function for classification models

custom_samplers

A formatted list of sampler functions

Value

A named list of samplers

Examples


# construct_samplers() can create a sequence of samplers using a fully-default 
# approach, a hybrid approach, or a fully-customized approach. All approaches
# require a roadmap and samplers. 

rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

# Fully-default approach

construct_samplers(
  roadmap = rm, 
  default_regression_sampler = sample_lm, 
  default_classification_sampler = sample_rpart
)

# Hybrid approach

construct_samplers(
  roadmap = rm, 
  default_regression_sampler = sample_lm,
  default_classification_sampler = sample_rpart,
  custom_samplers = list(
    list(vars = "hcovany", sampler = sample_rpart)
  )
)

# Fully-customized approach

construct_samplers(
  roadmap = rm, 
  custom_samplers = list(
    list(vars = c("hcovany", "empstat", "classwkr"), sampler = sample_rpart),
    list(vars = c("age", "famsize", "transit_time", "inctot"), sampler = sample_lm)
  )
)

Construct a list of tuning grids for hyperparameter tuning predictive models

Description

Construct a list of tuning grids for hyperparameter tuning predictive models

Usage

construct_tuners(
  roadmap,
  default_regression_tuner = NULL,
  default_classification_tuner = NULL,
  custom_tuners = NULL
)

Arguments

roadmap

A roadmap object

default_regression_tuner

A tuner.

default_classification_tuner

A tuner.

custom_tuners

A formatted list of tuners.

Value

A named list of tuners

Examples


# construct_tuners() can create a sequence of tuners using a fully-default 
# approach, a hybrid approach, or a fully-customized approach. All approaches
# require a roadmap and tuners. 

rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

tuner_reg <- list(
  v = 3,
  grid = 3,
  metrics = yardstick::metric_set(yardstick::rmse)
)

tuner_cat <- list(
  v = 3,
  grid = 3,
  metrics = yardstick::metric_set(yardstick::roc_auc)
)

# Fully-default approach

construct_tuners(
  roadmap = rm, 
  default_regression_tuner = tuner_reg,
  default_classification_tuner = tuner_cat
)

# Hybrid approach

tuner_cat2 <- list(
  v = 3,
  grid = 3,
  metrics = yardstick::metric_set(yardstick::precision)
)

construct_tuners(
  roadmap = rm, 
  default_regression_tuner = tuner_reg,
  default_classification_tuner = tuner_cat,
  custom_tuners = list(
    list(vars = "hcovany", tuner = tuner_cat2)
  )
)

# Fully-customized approach

construct_tuners(
  roadmap = rm, 
  custom_tuners = list(
    list(vars = c("hcovany", "empstat", "classwkr"), tuner = tuner_reg),
    list(vars = c("age", "famsize", "transit_time", "inctot"), tuner = tuner_cat)
  )
)

Convert `"NA"` values to `NA` for categorical variables

Description

Convert "NA" values to NA for categorical variables

Usage

convert_level_to_na(data)

Arguments

data

A data frame or tibble

Value

A data frame or tibble with "NA" converted to NA

Examples


data <- data.frame(
  x1 = c(1, 2, NA),  
  x2 = c("1", "2", "NA"),
  x3 = factor(c("1", "2", "NA")),
  x4 = factor(c("b", "NA", "a"), ordered = TRUE)
)

convert_level_to_na(data)

Convert `NA` values to `"NA"` for categorical variables

Description

Convert NA values to "NA" for categorical variables

Usage

convert_na_to_level(data)

Arguments

data

A data frame or tibble

Value

A data frame or tibble with NA converted to "NA"

Examples


data <- data.frame(
  x1 = c(1, 2, NA),
  x2 = c("1", "2", NA),
  x3 = factor(c("1", "2", NA)),
  x4 = factor(c("b", NA, "a"), levels = c("b", NA, "a"), ordered = TRUE)
)

convert_na_to_level(data)

Redefine `NA` value for a dataset.

Description

Redefine NA value for a dataset.

Usage

enforce_custom_na(data, col_schema)

Arguments

data

A data.frame object

col_schema

A col_schema from a schema object

Value

A data.frame

Examples


# create custom NA filter
example_na_custom <- example_na |>
  tidyr::replace_na(
    list("wages" = -999)
  )

example_na_expanded_custom <- enforce_custom_na(
  data = example_na_custom,
  col_schema = list(
    "wages" = list(
      dtype = "dbl",
      na_value = -999
     )
   )
 )

Add missing values where values should be missing according to _NA variables

Description

Add missing values where values should be missing according to _NA variables

Usage

enforce_na(data)

Arguments

data

A synthetic data frame with _NA columns

Value

A synthetic data frame with _NA columns that converts values that are labelled missing in an _NA variable to missing in the corresponding variable

Examples


example_na_expanded <- expand_na(data = example_na)

enforce_na(data = example_na_expanded)

Enforce a `roadmap`'s `schema` on its existing data

Description

Enforce a roadmap's schema on its existing data

Usage

enforce_schema(roadmap)

Arguments

roadmap

A roadmap object

Value

A roadmap object with modified conf_data, start_data, and schema information.

Examples

  
rm <- roadmap(conf_data = acs_conf, start_data = acs_start) |>
  update_schema(na_numeric_to_ind = TRUE)

enforce_schema(rm)

A df with different types of missingness

Description

A df with different types of missingness

Usage

example_na

Format

A tibble with 200 observations and 6 variables:

age: Age of respondent
sex: Sex of respondent with missingness at random
labor_force: Labor force status of respondent with structural missingness
hours: Hours work of respondent with missingness at random
wages: Wages earned with structural missingness

Add new variables that indicate if a value is "missing" or "not missing" for original variables that contain NA

Description

Add new variables that indicate if a value is "missing" or "not missing" for original variables that contain NA

Usage

expand_na(
  data,
  types = c("chr", "dbl", "fct", "lgl", "int", "ord"),
  skip_vars = NULL
)

Arguments

data

A data frame

types

A vector of variables types to expand

skip_vars

A character vector of variables that shouldn't be expanded

Value

An augmented data frame with the original variables and new variables that contain the missingness patterns of variables with NA

Examples


expand_na(data = example_na, type = c("dbl", "int"))

An S3 method for inverting a step

Description

An S3 method for inverting a step

Usage

invert(object, predictions, ...)

Arguments

object

A recipe after fitting a model

predictions

A data frame with .pred

...

Other arguments

Value

A tibble with inverted model-generated values

Examples


data <- tibble::tibble(
  y = rlnorm(n = 1000, meanlog = 0, sdlog = 1),
  x = rnorm(n = 1000)
)

adj <- recipes::recipe(y ~ x, data = data) |>
  recipes::step_BoxCox(recipes::all_outcomes()) |>
  recipes::prep()
  
invert(
  object = adj$steps[[1]], 
  predictions = tibble::tibble(.pred = adj[["template"]][["y"]])
)

Invert a Box-Cox transformation

Description

Invert a Box-Cox transformation

Usage

## S3 method for class 'step_BoxCox'
invert(object, predictions, ...)

Arguments

object

A recipe after fitting a model

predictions

A data frame with .pred

...

Other arguments

Value

A tibble with the Box-Cox transformation inverted for .pred

Examples


data <- tibble::tibble(
  y = rlnorm(n = 1000, meanlog = 0, sdlog = 1),
  x = rnorm(n = 1000)
)

adj <- recipes::recipe(y ~ x, data = data) |>
  recipes::step_BoxCox(recipes::all_outcomes()) |>
  recipes::prep()
  
invert(
  object = adj$steps[[1]], 
  predictions = tibble::tibble(.pred = adj[["template"]][["y"]])
)

Invert a Yeo-Johnson transformation

Description

Invert a Yeo-Johnson transformation

Usage

## S3 method for class 'step_YeoJohnson'
invert(object, predictions, ...)

Arguments

object

A recipe after fitting a model

predictions

A data frame with .pred

...

Other arguments

Value

A tibble with the Yeo_johnson transformation inverted for .pred

Examples


data <- tibble::tibble(
  y = rlnorm(n = 1000, meanlog = 0, sdlog = 1),
  x = rnorm(n = 1000)
)

adj <- recipes::recipe(y ~ x, data = data) |>
  recipes::step_YeoJohnson(recipes::all_outcomes()) |>
  recipes::prep()
  
invert(
  object = adj$steps[[1]], 
  predictions = tibble::tibble(.pred = adj[["template"]][["y"]])
)

Invert a log transformation

Description

Invert a log transformation

Usage

## S3 method for class 'step_log'
invert(object, predictions, ...)

Arguments

object

A recipe after fitting a model

predictions

A data frame with .pred

...

Other arguments

Value

A tibble with the log transformation inverted for .pred

Examples


data <- tibble::tibble(
  y = rlnorm(n = 1000, meanlog = 0, sdlog = 1),
  x = rnorm(n = 1000)
)

adj <- recipes::recipe(y ~ x, data = data) |>
  recipes::step_log(recipes::all_outcomes()) |>
  recipes::prep()
  
invert(
  object = adj$steps[[1]], 
  predictions = tibble::tibble(.pred = adj[["template"]][["y"]])
)

Kolmogorov-Smirnov distance

Description

Kolmogorov-Smirnov distance

Usage

ks_distance(data, ...)

## S3 method for class 'data.frame'
ks_distance(data, truth, estimate, na_rm = TRUE, case_weights = NULL, ...)

ks_distance_vec(truth, estimate, na_rm = TRUE, case_weights = NULL, ...)

Arguments

data

A data.frame containing the columns specified by the truth and estimate arguments.

...

Not currently used.

truth

The column identifier for the true results (that is numeric). This should be an unquoted column name although this argument is passed by expression and supports quasiquotation (you can unquote column names). For ⁠_vec()⁠ functions, a numeric vector.

estimate

The column identifier for the predicted results (that is also numeric). As with truth this can be specified different ways but the primary method is to use an unquoted variable name. For ⁠_vec()⁠ functions, a numeric vector.

na_rm

A logical value indicating whether NA values should be stripped before the computation proceeds.

case_weights

This is a placeholder for now and will be added when case_weights are added to tidysynthesis.

Value

For ks_distance_vec(), a single numeric value (or NA).

A single numeric value (or NA).

Examples


ks1 <- data.frame(x = 1:100, y = 101:200)

ks_distance(data = ks1, truth = x, estimate = y)


ks1 <- data.frame(x = 1:100, y = 101:200)

ks_distance(data = ks1, truth = x, estimate = y)


ks1 <- data.frame(x = 1:100, y = 101:200)

ks_distance_vec(truth = ks1$x, estimate = ks1$y)

Create a noise object

Description

Create a noise object

Usage

noise(add_noise = FALSE, mode = "regression", noise_func = NULL, ...)

Arguments

add_noise

Boolean, TRUE if adding noise

mode

String, one of "regression" or "classification"

noise_func

A function that adds noise to

...

Optional named additional arguments to pass to noise_func(...)

Value

A noise object

Examples


# create default noise object
noise()

# create noise object for classification
noise(
  add_noise = TRUE,
  mode = "classification",
  noise_func = add_noise_cat_unif
)

# create noise object for regression
noise(
  add_noise = TRUE,
  mode = "regression",
  noise_func = add_noise_kde,
  n_ntiles = 10
)

Create a presynth object

Description

Create a presynth object

Usage

presynth(roadmap, synth_spec)

Arguments

roadmap

A roadmap object from roadmap().

synth_spec

A synth_spec object from synth_spec().

Value

A presynth object.

Examples


# create roadmap
rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
) 

rpart_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

rpart_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "classification")

synth_spec1 <- synth_spec(
  default_regression_model = rpart_mod_reg,
  default_regression_sampler = sample_rpart,
  default_classification_model = rpart_mod_class,
  default_classification_sampler = sample_rpart
)

# create a presynth object
# use defaults for noise, constraints, and replicates
presynth(
  roadmap = rm,
  synth_spec = synth_spec1
)

Print the constraints object to the console with formatting

Description

Print the constraints object to the console with formatting

Usage

## S3 method for class 'constraints'
print(x, ...)

Arguments

x

A constraints object

...

further arguments passed to or from other methods (not currently used).

Value

Invisibly returns the input constraints object.

Examples


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

constraints_df_num <- 
  tibble::tribble(~var, ~min, ~max, ~conditions,
                  "transit_time", 0, 300, "TRUE")

constraints <- constraints(
  schema = rm[["schema"]],
  constraints_df_num = constraints_df_num,
  max_z_num = 0
)

print(constraints)

Print the noise object to the console with formatting

Description

Print the noise object to the console with formatting

Usage

## S3 method for class 'noise'
print(x, ...)

Arguments

x

A noise object

...

further arguments passed to or from other methods (not currently used).

Value

Invisibly returns the input noise object.

Examples


print(noise())

Print the postsynth object to the console with formatting

Description

Print the postsynth object to the console with formatting

Usage

## S3 method for class 'postsynth'
print(x, ...)

Arguments

x

A postsynth object

...

further arguments passed to or from other methods (not currently used).

Value

Invisibly returns the input postsynth object.

Examples


# create roadmap
rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
) 

rpart_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

rpart_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "classification")

synth_spec1 <- synth_spec(
  default_regression_model = rpart_mod_reg,
  default_regression_sampler = sample_rpart,
  default_classification_model = rpart_mod_class,
  default_classification_sampler = sample_rpart
)

# create a presynth object
# use defaults for noise, constraints, and replicates
presynth1 <- presynth(
  roadmap = rm,
  synth_spec = synth_spec1
)

# synthesize!
set.seed(1)
postsynth1 <- synthesize(presynth = presynth1)

print(postsynth1)

print method for presynth objects

Description

print method for presynth objects

Usage

## S3 method for class 'presynth'
print(x, ...)

Arguments

x

A presynth object

...

further arguments passed to or from other methods (not currently used).

Value

A presynth object

Examples


# create roadmap
rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
) 

rpart_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

rpart_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "classification")

synth_spec1 <- synth_spec(
  default_regression_model = rpart_mod_reg,
  default_regression_sampler = sample_rpart,
  default_classification_model = rpart_mod_class,
  default_classification_sampler = sample_rpart
)

# create a presynth object
# use defaults for noise, constraints, and replicates
presynth <- presynth(
  roadmap = rm,
  synth_spec = synth_spec1
)

print(presynth)

Print the replicates object to the console with formatting

Description

Print the replicates object to the console with formatting

Usage

## S3 method for class 'replicates'
print(x, ...)

Arguments

x

A replicates object

...

further arguments passed to or from other methods (not currently used).

Value

Invisibly returns the input replicates object.

Examples


rep <- replicates(
  start_data_replicates = 2,
  model_sample_replicates = 2,
  end_to_end_replicates = 2
)
  
print(rep)

Print the schema object to the console with formatting

Description

Print the schema object to the console with formatting

Usage

## S3 method for class 'schema'
print(x, ...)

Arguments

x

A schema object

...

further arguments passed to or from other methods (not currently used).

Value

Invisibly returns the input schema object.

Examples


# default inferred schema
schema1 <- schema(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

print(schema1)

Print the start_method object to the console with formatting

Description

Print the start_method object to the console with formatting

Usage

## S3 method for class 'start_method'
print(x, ...)

Arguments

x

A start_method object

...

further arguments passed to or from other methods (not currently used).

Value

A start_method object

Examples

 

print(start_method())

Print the replicates object to the console with formatting

Description

Print the replicates object to the console with formatting

Usage

## S3 method for class 'synth_spec'
print(x, ...)

Arguments

x

A replicates object

...

further arguments passed to or from other methods (not currently used).

Value

A synth_spec object

Examples


synth_spec <- synth_spec()

print(synth_spec)

Print method for `visit_sequence` objects

Description

Print method for visit_sequence objects

Usage

## S3 method for class 'visit_sequence'
print(x, ...)

Arguments

x

A visit_sequence object

...

further arguments passed to or from other methods (not currently used).

Value

Invisibly returns the input visit_sequence object.

Examples


rm <- roadmap(
  conf_data = acs_conf_nw, 
  start_data = acs_start_nw
)

print(rm[["visit_sequence"]])

Create a replicates object

Description

Create a replicates object

Usage

replicates(
  start_data_replicates = 1,
  model_sample_replicates = 1,
  end_to_end_replicates = 1
)

Arguments

start_data_replicates

The number of starting data replicates to use. Note that if no start_method is provided, all start data replicates will be identical.

model_sample_replicates

The number of replicates for the conditional modeling process, including modeling and sampling new synthetic values.

end_to_end_replicates

The number of replicates for the entire synthesis process, including all previously specified steps.

Value

A new replicates object.

Examples


replicates(
  start_data_replicates = 2,
  model_sample_replicates = 2,
  end_to_end_replicates = 2
)

Add, update, or reset a `replicates` object within an existing `roadmap`.

Description

Add, update, or reset a replicates object within an existing roadmap.

Usage

add_replicates(roadmap, replicates)

update_replicates(roadmap, ...)

reset_replicates(roadmap)

Arguments

roadmap

A roadmap object

replicates

A replicates object.

...

Optional named parameters passed to replicates().

Value

A new roadmap object.

A new roadmap object with the added replicates.

A new roadmap object with updated replicates.

A new roadmap object with reset replicates.

Examples


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

new_replicates <- replicates(end_to_end_replicates = 2)

rm |>
  add_replicates(new_replicates)


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

rm |>
  update_replicates(start_data_replicates = 3)


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

rm <- rm |>
  add_replicates(replicates(start_data_replicates = 3))
  
 reset_replicates(roadmap = rm)

Create a roadmap

Description

A roadmap is a container object that aggregates information required to specify the order of operations for synthesis modeling and sampling steps.

Usage

roadmap(
  conf_data,
  start_data,
  start_method = NULL,
  schema = NULL,
  visit_sequence = NULL,
  replicates = NULL,
  constraints = NULL
)

Arguments

conf_data

A data.frame of confidential data.

start_data

A data.frame of starting data used to initialize the process.

start_method

An optional start_method object.

schema

An optional schema object.

visit_sequence

An optional visit_sequence object.

replicates

An optional replicates object.

constraints

An optional constraints object.

Details

Users initiate a roadmap object with conf_data and start_data. All other objects will either be completed with defaults or specified interactively via the provided API.

Value

A new roadmap object.

Examples

roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw,
  start_method = start_method(
    start_func = start_resample, n = 1000
  )
)

Sample the conditional distribution created by a generalized linear model

Description

Currently, logistic and poisson regression are supported using parsnip and the standard glm engine. Note that poisson regression requires the suggested poissonreg library.

Usage

sample_glm(model, new_data, conf_data)

Arguments

model

A "model_fit" object created by parsnip

new_data

A data frame with predictors

conf_data

A data frame with original confidential predictors

Value

A numeric vector of predictions

Examples


acs_conf <- acs_conf |>
  tidyr::drop_na()

logistic_mod <- parsnip::logistic_reg() |>
  parsnip::set_engine("glm") |>
  parsnip::set_mode(mode = "classification")

classification_rec <- recipes::recipe(hcovany ~ ., data = acs_conf)

model_class <- workflows::workflow() |>
  workflows::add_model(spec = logistic_mod) |>
  workflows::add_recipe(recipe = classification_rec) |>
  parsnip::fit(data = acs_conf)

set.seed(1)
sample1 <- sample_glm(
  model = model_class, 
  new_data = acs_conf[1:3, ], 
  conf_data = acs_conf
)

Sample the conditional distribution created by a linear model

Description

Sample the conditional distribution created by a linear model

Usage

sample_lm(model, new_data, conf_data)

Arguments

model

A "model_fit" object created by parsnip::linear_reg()

new_data

A data frame with predictors

conf_data

A data frame with original confidential predictors

Value

A numeric vector of predictions

Examples


lm_mod <- parsnip::linear_reg() |>
  parsnip::set_engine("lm") |>
  parsnip::set_mode(mode = "regression")

regression_rec <- recipes::recipe(inctot ~ ., data = acs_conf)

model_reg <- workflows::workflow() |>
  workflows::add_model(spec = lm_mod) |>
  workflows::add_recipe(recipe = regression_rec) |>
  parsnip::fit(data = acs_conf)

set.seed(1)
sample1 <- sample_lm(
  model = model_reg, 
  new_data = acs_conf[1:3, ], 
  conf_data = acs_conf
)

Sample the conditional distribution created by a ranger rf model

Description

Sample the conditional distribution created by a ranger rf model

Usage

sample_ranger(model, new_data, conf_data)

Arguments

model

A "model_fit" object created by parsnip::ranger()

new_data

A data frame with predictors

conf_data

A data frame with original confidential predictors

Value

A numeric vector of predictions

Examples


rf_mod_regression <- parsnip::rand_forest(trees = 500, min_n = 1) |>
  parsnip::set_engine(engine = "ranger") |>
  parsnip::set_mode(mode = "regression") |>
  parsnip::set_args(quantreg = TRUE)

regression_rec <- recipes::recipe(age ~ ., data = acs_conf)

model_reg <- workflows::workflow() |>
  workflows::add_model(spec = rf_mod_regression) |>
  workflows::add_recipe(recipe = regression_rec) |>
  parsnip::fit(data = acs_conf)

set.seed(1)
sample1 <- sample_ranger(
  model = model_reg, 
  new_data = acs_conf[1:3, ], 
  conf_data = acs_conf
)

Sample the conditional distribution created by a CART model

Description

Sample the conditional distribution created by a CART model

Usage

sample_rpart(model, new_data, conf_data, ignore_zeros = TRUE)

Arguments

model

A "model_fit" object created by rpart

new_data

A data frame with predictors

conf_data

A data frame with original confidential predictors

ignore_zeros

Should a vector of all 0 observations return NA for the l-diversity calculation. Defaults to TRUE.

Value

A numeric vector of predictions

Examples


rpart_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine("rpart") |>
  parsnip::set_mode(mode = "regression")

 regression_rec <- recipes::recipe(inctot ~ ., data = acs_conf)
 
 model_reg <- workflows::workflow() |>
   workflows::add_model(spec = rpart_mod_reg) |>
   workflows::add_recipe(recipe = regression_rec) |>
   parsnip::fit(data = acs_conf)
 
 set.seed(1)
 sample1 <- sample_rpart(
   model = model_reg, 
   new_data = acs_conf[1:3, ], 
   conf_data = acs_conf
 )
 

rpart_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine("rpart") |>
  parsnip::set_mode(mode = "classification")

classification_rec <- recipes::recipe(hcovany ~ ., data = acs_conf)

model_reg <- workflows::workflow() |>
  workflows::add_model(spec = rpart_mod_class) |>
  workflows::add_recipe(recipe = classification_rec) |>
  parsnip::fit(data = acs_conf)

set.seed(1)
sample1 <- sample_rpart(
  model = model_reg, 
  new_data = acs_conf[1:10, ], 
  conf_data = acs_conf
)

Generate a `schema` object.

Description

Generate a schema object.

Usage

schema(
  conf_data,
  start_data,
  col_schema = NULL,
  enforce = TRUE,
  coerce_to_factors = FALSE,
  coerce_to_doubles = FALSE,
  na_factor_to_level = TRUE,
  na_numeric_to_ind = TRUE
)

Arguments

conf_data

A data frame to be synthesized.

start_data

A data frame with starting variables.

col_schema

An optional named list of columns in the confidential data with their properties, including data type and factor levels. If NULL or only partially specified, col_schema will be inferred from the confidential data. See example code for formatting.

enforce

Boolean that if true, will preprocess both conf_data and start_data to align with col_schema and the arguments below.

coerce_to_factors

Boolean that if true, coerces categorical data types (chr, fct, ord) to base R factors when enforce_schema is called.

coerce_to_doubles

Boolean that if true, coerces columns specified as dbl in col_schema to base R doubles when enforce_schema is called.

na_factor_to_level

Boolean that if true, applies convert_level_to_na() to factor variables when enforce_schema is called.

na_numeric_to_ind

Boolean that if true, applies expand_na() to numeric data to create logical missingness indicators when enforce_schema is called.

Value

A schema object.

Examples


conf_data <- data.frame(
  var1 = c("1", "1", "2"),
  var2 = c(1L, 2L, 3L),
  var3 = c(1.1, 2.2, 3.3)
)

start_data <- dplyr::select(conf_data, var1)

# default inferred schema
schema(
  conf_data = conf_data,
  start_data = start_data
)

# overwriting factor levels
schema(
  conf_data = conf_data,
  start_data = start_data,
  col_schema = list(
    "var1" = list(
      "dtype" = "fct",
      "levels" = c("1", "2", "3")
    )
  ),
  coerce_to_factors = TRUE
)

Add, update, or reset a `schema` object within an existing `roadmap`.

Description

Add, update, or reset a schema object within an existing roadmap.

Usage

add_schema(roadmap, schema)

update_schema(roadmap, ...)

reset_schema(roadmap)

Arguments

roadmap

A roadmap object

schema

A schema object.

...

Optional named parameters passed to schema().

Value

A new roadmap object.

A roadmap object with added schema.

A roadmap object with updated schema.

A roadmap object with reset schema.

Examples


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

acs_schema <- schema(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw,
  na_numeric_to_ind = TRUE
)

rm |>
  add_schema(schema = acs_schema)


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

rm |>
  update_schema(na_numeric_to_ind = TRUE)


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
)

rm <- rm |>
  update_schema(na_numeric_to_ind = TRUE)

reset_schema(roadmap = rm)

Create a `start_method` object.

Description

A start_method gets executed prior to running a synthesis. This modifies the start_data, typically randomly, to provide greater disclosure risk protections.

Usage

start_method(start_func = NULL, ...)

Arguments

start_func

A function that accepts and returns a data.frame. If none provided .identity_start() is used.

...

Optional keyword arguments passed to start_func(...)

Value

A start_method object

Examples


# basic usage
start_method(start_func = start_resample)

# adjust the number of observations
start_method(
  start_func = start_resample,
  start_data = acs_start_nw,
  n = 10
)

# adjust the number of observations and use all combinations as support
start_method(
  start_func = start_resample,
  start_data = acs_start_nw,
  n = 10, 
  inv_noise_scale = 1,
  support = "all"
)

Add, update, or reset a start method within an existing `roadmap`.

Description

Add, update, or reset a start method within an existing roadmap.

Usage

add_start_method(roadmap, start_method)

update_start_method(roadmap, ...)

remove_start_method(roadmap)

Arguments

roadmap

A roadmap object

start_method

A start_method object.

...

Optional named parameters passed to start_method()

Value

A new roadmap object.

A new roadmap object with added start_method.

A new roadmap object with updated start_method.

A new roadmap object with removed start_method.

Examples


rm <- roadmap( 
 conf_data = acs_conf_nw,
 start_data = acs_start_nw,
)

add_start_method(
  roadmap = rm,
  start_method = start_method()
)


rm <- roadmap( 
 conf_data = acs_conf_nw,
 start_data = acs_start_nw
)

update_start_method(
  roadmap = rm,
  start_method = start_method()
)

rm <- roadmap( 
 conf_data = acs_conf_nw,
 start_data = acs_start_nw,
 start_method = start_method()
)

remove_start_method(
  roadmap = rm
)

Specify a resampling scheme for start_data

Description

Specify a resampling scheme for start_data

Usage

start_resample(
  start_data,
  n = NULL,
  inv_noise_scale = NULL,
  support = c("observed", "all")
)

Arguments

start_data

A data.frame

n

An optional integer sample size. If unspecified, n = nrow(start_data)

inv_noise_scale

An optional parameter to set randomized noise to the proportions of records with different start_data characteristics. Corresponds to a privacy loss budget under epsilon differential privacy.

support

A string that specifies the method of resampling from the start_data domain.

Value

A start_method object for resampling starting data

Examples

 
 
start_method(
  start_func = start_resample, n = 1000
)

Create a `synth_spec` object

Description

The synth_spec object holds specifications for modeling and sampling components for sequential synthetic data generation. Each component has an associated ⁠construct_*⁠ function called when creating a presynth object.

Usage

synth_spec(
  default_regression_model = NULL,
  default_classification_model = NULL,
  custom_models = NULL,
  default_regression_steps = NULL,
  default_classification_steps = NULL,
  custom_steps = NULL,
  default_regression_sampler = NULL,
  default_classification_sampler = NULL,
  custom_samplers = NULL,
  default_regression_noise = NULL,
  default_classification_noise = NULL,
  custom_noise = NULL,
  default_regression_tuner = NULL,
  default_classification_tuner = NULL,
  custom_tuners = NULL,
  default_extractor = NULL,
  custom_extractors = NULL,
  invert_transformations = TRUE,
  enforce_na = TRUE
)

Arguments

default_regression_model

A model_spec object from library(parsnip) for use in regression models.

default_classification_model

A model_spec object from library(parsnip) for use in classification models.

custom_models

A list of named lists each with two elements: vars for variable names, and model for their associated model. from library(parsnip).

default_regression_steps

A list of recipe::step_ function(s) from library(recipes) for use in regression models.

default_classification_steps

A list of recipe::step_ function(s) from library(recipes) for use in classification models.

custom_steps

A list of named lists each with two elements: vars for variable names, and steps for their associated recipe.

default_regression_sampler

A sampling function for drawing new values from regression models.

default_classification_sampler

A sampling function for drawing new values from classification models.

custom_samplers

A list of named lists each with two elements: vars for variable names, and sampler for their associated sampler

default_regression_noise

A noise function for adding noise to numeric values.

default_classification_noise

A noise function for adding noise to classification values.

custom_noise

A list of named lists each with two elements: vars for variable names, and noise for their associated noise

default_regression_tuner

A tuner from library(tune) for use in regression models.

default_classification_tuner

A tuner from library(tune) for use in classification models.

custom_tuners

A list of named lists each with two elements: vars for variable names, and tuner for their associated tuner

default_extractor

An optional method for extracting workflows or extracts from workflows.

custom_extractors

A list of named lists each with two elements: vars for variable names, and extractor for their associated extractor

invert_transformations

A Boolean for if outcome variable transformations applied through recipes should be inverted during synthesis. recipes need ids that begin with "outcome".

enforce_na

A Boolean for if NA values should be added into the synthetic data with enforce_na() during synthesis. An alternative approach is to add the NA values after synthesis

Value

A synth_spec object

Examples


rpart_mod <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

lm_mod <- parsnip::linear_reg() |> 
  parsnip::set_engine("lm") |>
  parsnip::set_mode(mode = "regression")

step1 <- function(x) {
 x |>
   recipes::step_center(recipes::all_predictors(), id = "center")
}

step2 <- function(x) {
  x |>
    recipes::step_scale(recipes::all_predictors(), id = "scale")
}

step3 <- function(x) { x |> step1() |> step2() }


synth_spec(
 default_regression_model = rpart_mod,
 custom_models = list(
   list("vars" = c("var1", "var2"), 
        "model" = lm_mod)
 ),
 default_regression_steps = step1,
 custom_steps = list(
   list("vars" = c("var2", "var3"),
        "steps" = step2),
   list("vars" = c("var4"), 
        "steps" = step3)
 ),
 default_regression_sampler = sample_rpart,
 custom_samplers = list(
   list("vars" = c("var1", "var2"), 
        "sampler" = sample_lm)
 )
)

Add, update, or remove extractors from a `synth_spec` object

Description

Add, update, or remove extractors from a synth_spec object

Usage

add_custom_extractors(synth_spec, ...)

update_custom_extractors(synth_spec, ...)

remove_custom_extractors(synth_spec)

Arguments

synth_spec

A synth_spec object

...

Optional named lists with two elements, vars and extractor, mapping variable names to extractors.

Value

A new synth_spec object.

A new synth_spec object with added custom extractors.

A new synth_spec object with updated custom extractors.

A new synth_spec object with removed custom extractors.

Examples


synth_spec <- synth_spec()

add_custom_extractors(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "extractor" = parsnip::extract_fit_engine)
)


synth_spec <- synth_spec()

update_custom_extractors(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "extractor" = parsnip::extract_fit_engine)
)


synth_spec <- synth_spec()

synth_spec <- add_custom_extractors(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "extractor" = parsnip::extract_fit_engine)
)

remove_custom_extractors(synth_spec = synth_spec)

Inspections for `synth_spec` components

Description

Inspections for synth_spec components

Arguments

z

Object

Value

Boolean if matches class type

Add, update, or remove custom models from a `synth_spec` object

Description

Add, update, or remove custom models from a synth_spec object

Usage

add_custom_models(synth_spec, ...)

update_custom_models(synth_spec, ...)

remove_custom_models(synth_spec)

Arguments

synth_spec

A synth_spec object

...

Optional named lists with two elements, vars and model, mapping variable names to model_spec objects from library(parsnip).

Value

A new synth_spec object.

A new synth_spec object with added custom models.

A new synth_spec object with updated custom models.

A new synth_spec object with removed custom models.

Examples


synth_spec <- synth_spec()

dt_reg_mod <- parsnip::decision_tree() |>
  parsnip::set_engine("rpart") |>
  parsnip::set_mode("regression")

add_custom_models(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "model" = dt_reg_mod)
)


synth_spec <- synth_spec()

dt_reg_mod <- parsnip::decision_tree() |>
  parsnip::set_engine("rpart") |>
  parsnip::set_mode("regression")

update_custom_models(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "model" = dt_reg_mod)
)


synth_spec <- synth_spec()

dt_reg_mod <- parsnip::decision_tree() |>
  parsnip::set_engine("rpart") |>
  parsnip::set_mode("regression")

synth_spec <- update_custom_models(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "model" = dt_reg_mod)
)

remove_custom_models(synth_spec = synth_spec)

Add, update, or remove noise from a `synth_spec` object

Description

Add, update, or remove noise from a synth_spec object

Usage

add_custom_noise(synth_spec, ...)

Arguments

synth_spec

A synth_spec object

...

Optional named lists with two elements, vars and noise, mapping variable names to samplers.

Value

A new synth_spec object.

A new synth_spec object with added custom noise.

Examples


synth_spec <- synth_spec()

noise1 <- noise(
  add_noise = TRUE, 
  noise_func = add_noise_kde,
  noise_params = list(
    n_ntiles = 2
  )
)

add_custom_noise(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "noise" = noise1)
)

Add, update, or remove recipe recipes from a `synth_spec` object

Description

Add, update, or remove recipe recipes from a synth_spec object

Usage

add_custom_steps(synth_spec, ...)

update_custom_steps(synth_spec, ...)

remove_custom_steps(synth_spec)

Arguments

synth_spec

A synth_spec object

...

Optional named arguments mapping variables to lists of recipe::recipe_ function(s) from library(recipes).

Value

A new synth_spec object.

A new synth_spec object with added custom steps.

A new synth_spec object with updated custom steps.

A new synth_spec object with removed custom steps.

Examples


synth_spec <- synth_spec()

step1 <- function(x) {
  x |> recipes::step_center(recipes::all_predictors(), id = "center")
}

add_custom_steps(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "steps" = step1)
)


synth_spec <- synth_spec()

step1 <- function(x) {
  x |> recipes::step_center(recipes::all_predictors(), id = "center")
}

update_custom_steps(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "steps" = step1)
)


synth_spec <- synth_spec()

step1 <- function(x) {
  x |> recipes::step_center(recipes::all_predictors(), id = "center")
}

synth_spec <- add_custom_steps(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "steps" = step1)
)

remove_custom_steps(synth_spec = synth_spec)

Add, update, or remove samplers from a `synth_spec` object

Description

Add, update, or remove samplers from a synth_spec object

Usage

add_custom_samplers(synth_spec, ...)

update_custom_samplers(synth_spec, ...)

remove_custom_samplers(synth_spec)

update_custom_noise(synth_spec, ...)

remove_custom_noise(synth_spec)

Arguments

synth_spec

A synth_spec object

...

Optional named lists with two elements, vars and sampler, mapping variable names to samplers.

Value

A new synth_spec object.

A new synth_spec object with added custom samplers.

A new synth_spec object with updated custom samplers.

A new synth_spec object with removed custom samplers.

A new synth_spec object with updated custom noise.

A new synth_spec object with removed custom noise.

Examples


synth_spec <- synth_spec()

add_custom_samplers(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "sampler" = sample_rpart)
)


synth_spec <- synth_spec()

update_custom_samplers(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "sampler" = sample_rpart)
)


synth_spec <- synth_spec()

synth_spec <- add_custom_samplers(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "sampler" = sample_rpart)
)

remove_custom_samplers(synth_spec = synth_spec)


synth_spec <- synth_spec()

noise1 <- noise(
  add_noise = TRUE, 
  noise_func = add_noise_kde,
  noise_params = list(
    n_ntiles = 2
  )
)

update_custom_noise(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "noise" = noise1)
)


synth_spec <- synth_spec()

noise1 <- noise(
  add_noise = TRUE, 
  noise_func = add_noise_kde,
  noise_params = list(
    n_ntiles = 2
  )
)

synth_spec <- add_custom_noise(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "noise" = noise1)
)

remove_custom_noise(synth_spec = synth_spec)

Add, update, or remove tuners from a `synth_spec` object

Description

Add, update, or remove tuners from a synth_spec object

Usage

add_custom_tuners(synth_spec, ...)

update_custom_tuners(synth_spec, ...)

remove_custom_tuners(synth_spec)

Arguments

synth_spec

A synth_spec object

...

Optional named lists with two elements, vars and tuner, mapping variable names to tuners.

Value

A new synth_spec object.

A new synth_spec object with added custom tuners.

A new synth_spec object with updated custom tuners.

A new synth_spec object with removed custom tuners.

Examples


synth_spec <- synth_spec()

tuner1 <- list(
  v = 3,
  grid = 3,
  metrics = yardstick::metric_set(yardstick::rmse)
)

add_custom_tuners(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "tuner" = tuner1)
)


synth_spec <- synth_spec()

tuner1 <- list(
  v = 3,
  grid = 3,
  metrics = yardstick::metric_set(yardstick::rmse)
)

update_custom_tuners(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "tuner" = tuner1)
)


synth_spec <- synth_spec()

tuner1 <- list(
  v = 3,
  grid = 3,
  metrics = yardstick::metric_set(yardstick::rmse)
)

synth_spec <- add_custom_tuners(
  synth_spec = synth_spec, 
  list("vars" = c("a", "b", "c"), "tuner" = tuner1)
)

remove_custom_tuners(synth_spec = synth_spec)

Synthesize a data set

Description

Synthesize a data set

Usage

synthesize(presynth, progress = FALSE)

Arguments

presynth

A presynth object created by presynth().

progress

A single logical. Should a progress be displayed?

Value

A postsynth object.

Examples


# create roadmap
rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
) 

rpart_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

rpart_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "classification")

synth_spec1 <- synth_spec(
  default_regression_model = rpart_mod_reg,
  default_regression_sampler = sample_rpart,
  default_classification_model = rpart_mod_class,
  default_classification_sampler = sample_rpart
)

# create a presynth object
# use defaults for noise, constraints, and replicates
presynth1 <- presynth(
  roadmap = rm,
  synth_spec = synth_spec1
)

# synthesize!
set.seed(1)
postsynth1 <- synthesize(presynth = presynth1)

Generate syntheses from multiple `presynth` objects.

Description

Generate syntheses from multiple presynth objects.

Usage

tune_synthesis(
  presynths,
  postprocessing_func,
  metadata_func = NULL,
  simplify_post = FALSE,
  seed = NULL
)

Arguments

presynths

A list of presynth objects

postprocessing_func

A function with arguments "synth_id", "synth_name", and "postsynth" that performs any desired postprocessing operations, like writing

metadata_func

An optional function with argument "presynth" that extracts specified information from each presynth object and returns a list. Each list element becomes an additional column in the output metadata.

simplify_post

Boolean that, if true, expects postprocessing_func to return a list corresponding to the row of the output dataframe (one per synthesis).

seed

A RNG seed to pass to set.seed()

Value

A post_tunesynth object.

Examples


rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
) 

dt_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

dt_mod_reg_cp <- parsnip::decision_tree(cost_complexity = 0.01) |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

dt_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "classification")

# synth specs
synth_spec1 <- synth_spec(
  default_regression_model = dt_mod_reg,
  default_regression_sampler = sample_rpart,
  default_classification_model = dt_mod_class,
  default_classification_sampler = sample_rpart
)

synth_spec2 <- synth_spec(
  default_regression_model = dt_mod_reg_cp,
  default_regression_sampler = sample_rpart,
  default_classification_model = dt_mod_class,
  default_classification_sampler = sample_rpart
)


presynth1 <- presynth(
  roadmap = rm,
  synth_spec = synth_spec1
)

presynth2 <- presynth(
  roadmap = rm,
  synth_spec = synth_spec2
)

postproc_f_null <- function(synth_id, synth_name, postsynth) {
  return(postsynth[["synthetic_data"]])
}

tune_synthesis(
  presynths = list(presynth1, presynth2),
  postprocessing_func = postproc_f_null,
  seed = 12345
)

Update `presynth` object

Description

Update presynth object

Usage

update_presynth(presynth, roadmap = NULL, synth_spec = NULL)

Arguments

presynth

A presynth object

roadmap

An optional roadmap object

synth_spec

An optional synth_spec object

Value

A presynth object.

Examples


# create roadmap
rm <- roadmap(
  conf_data = acs_conf_nw,
  start_data = acs_start_nw
) 

rpart_mod_reg <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "regression")

rpart_mod_class <- parsnip::decision_tree() |>
  parsnip::set_engine(engine = "rpart") |>
  parsnip::set_mode(mode = "classification")

synth_spec1 <- synth_spec(
  default_regression_model = rpart_mod_reg,
  default_regression_sampler = sample_rpart,
  default_classification_model = rpart_mod_class,
  default_classification_sampler = sample_rpart
)

# create a presynth object
# use defaults for noise, constraints, and replicates
presynth <- presynth(
  roadmap = rm,
  synth_spec = synth_spec1
)

lm_mod <- parsnip::linear_reg() |>
  parsnip::set_engine(engine = "lm") |>
  parsnip::set_mode(mode = "regression")

synth_spec2 <- synth_spec(
  default_regression_model = lm_mod,
  default_regression_sampler = sample_lm,
  default_classification_model = rpart_mod_class,
  default_classification_sampler = sample_rpart
)

Tidy API calls ———————————————————-

Description

Update non-custom synth_spec arguments

Usage

update_synth_spec(synth_spec, ...)

Arguments

synth_spec

A synth_spec object

...

Optional named keywords in synth_spec, with the exception of any ⁠custom_*⁠arguments

Value

A synth_spec

Examples


synth_spec <- synth_spec()

lm_mod <- parsnip::linear_reg() |> 
  parsnip::set_engine("lm") |>
  parsnip::set_mode(mode = "regression")

update_synth_spec(
  synth_spec, 
  default_regression_model = lm_mod
)

Generate a visit sequence.

Description

Generate a visit sequence.

Usage

visit_sequence(schema, weight_var = NULL, synthesize_weight = TRUE)

Arguments

schema

A schema object.

weight_var

A numeric weight for the weighted total ordering.

synthesize_weight

Boolean for if weight_var should be included in the visit sequence.

Value

A visit_sequence object.

Examples


df <- data.frame(
  factor_var = c("1", "1", "2"),
  vara = c(10000, 20000, 100000),
  varb = c(300, 200, 100),
  var_loss = c(1999999, 0, -1000000),
  weight = c(1000, 1000, 2000)
)

start_df <- dplyr::select(df, factor_var)

schema1 <- schema(
  conf_data = dplyr::select(df, -weight),
  start_data = start_df
)
                  
vs1 <- visit_sequence(
  schema = schema1
)

schema2 <- schema(
  conf_data = df,
  start_data = start_df
)

vs2 <- visit_sequence(
  schema = schema2,
  weight_var = weight,
  synthesize_weight = TRUE
)

Add or reset a `visit_sequence` object within an existing `roadmap`.

Description

Add or reset a visit_sequence object within an existing roadmap.

Usage

update_visit_sequence(roadmap, ...)

reset_visit_sequence(roadmap)

Arguments

roadmap

A roadmap object

...

Optional additional parameters.

Value

A new roadmap object.

A roadmap with an updated visit_sequence.

A new roadmap object with reset visit_sequence.

Examples


rm <- roadmap(
  conf_data = acs_conf_nw, 
  start_data = acs_start_nw
) 

rm |>
  update_visit_sequence(
    weight_var = wgt,
    synthesize_weight = TRUE
  )


rm <- roadmap(
  conf_data = acs_conf_nw, 
  start_data = acs_start_nw
) 

rm <- rm |>
  update_visit_sequence(
    weight_var = wgt,
    synthesize_weight = TRUE
  )

reset_visit_sequence(roadmap = rm)