saeplus_modelunitlevel: Estimate the unit level model

In SSA-Statistical-Team-Projects/SAEplus: Package for Geospatial Analysis & Small Area Estimation Post-Diagnostics

Estimate the unit level model

Description

This function prepares remote sensing data for imputation by checking the admin level shapefiles for geometric consistency. Prior to the imputation, an algorithm is used to select an appropriate set of variables that best predicts the outcome variable. The household survey data is used to create a synthetic census of households. The empirical best predictor is estimated using the household survey data and is used as the basis for imputation into the synthetic census.

Usage

saeplus_modelunitlevel(
  hhsurvey_dt,
  hhid_var,
  size_hh,
  adminshp_dt,
  target_id,
  geopolycensus_dt,
  geopopvar,
  geopoly_id = "id",
  crs_set = rep(4326, 3),
  agr_set = rep("constant", 3),
  cand_vars,
  cons_var,
  wgt_vartype = "hh",
  weight,
  create_dummy = TRUE,
  dummy_var,
  aggregate_id,
  ...,
  ncpu = 30,
  pline = 5006362,
  pline_transform = "inclusion_line",
  result_dir = getwd()
)

Arguments

hhsurvey_dt	object of class data.frame/data.table corresponding to household survey data (unit level data)
hhid_var	the household ID variable from the hhsurvey_dt object
size_hh	an integer/numeric for household size variable within the hhsurvey_dt object
adminshp_dt	an object of class sf, data.table and/or data.frame containing administrative level boundaries with multipolygons/polygons geometries
target_id	a character string representing an integer column vector for the admin level at which small area estimates will be computed for the poverty map
geopolycensus_dt	an object of class sf, data.table and/or data.frame containing polygon/multipolygon geometries and geospatial indicators
geopopvar	a character string for the population count variable name in the geopolycensus_dt
geopoly_id	a string/character variable representing the polygon ID within geopolycensus_dt
crs_set	an integer list of the coordinate reference systems for of the aforementioned objects i.e. the CRS for hhsurvey_dt, adminshp_dt and geopolycensus_dt in this order
agr_set	a character/string list representing attribute-geometry-relationships specified for each non-geometry attribute column and how it relates to the geometry, and can have one of following values "constant", "aggregate" and "identity". The default is constant. See details for more. The AGR will listed in the same order as the crs_set.
cand_vars	a character vector of candidate explanatory variables to be included in the model selection process
cons_var	the dependent variable for small area estimation (typically household per capita consumption)
wgt_vartype	a character string representing the weighting type. The options could be "hh", "pop" i.e. households vs population weights.
weight	a numeric/integer weight variable
create_dummy	if TRUE, a dummy variable will be created if dummy_var is specified. Variable must be from adminshp_dt
dummy_var	a list of variables from which a dummies will be created for each level.
aggregate_id	if argument is not NULL, all cand_vars will be aggregated at the admin level specified as the aggregate_id argument
...	include any set of arguments available within emdi::ebp() function as you see fit.
ncpu	the number of CPUs for parallelizing the small area estimation algorithm
pline	the national poverty line
pline_transform	select an order norm transformation method. There are three options. The default inclusion_line" option adds the poverty line value stipulated to the vector of welfare vector provided. The order norm value obtained from the normalization process is the converted value. The "interpolation_line" option applies a linear interpolation to outcome_var to estimate a conversion The "limsup_line" option takes the converted value of the greatest welfare value below the poverty line
result_dir	local folder directory where all tables and charts will be stored. If result_dir is not specified, the default is the result of getwd()
hhsurvey_lat	the latitude variable within the hh_dt object
hhsurvey_lon	the longitude variable within the hh_dt object

SSA-Statistical-Team-Projects/SAEplus documentation built on Aug. 24, 2022, 11:26 a.m.