In FPRgroup/FPEM: FPEM

Table of Contents

1. Introduction
2. Run models
3. Post-process
4. Plot model results

Introduction

In this vignette we will run two models, one for women in union and another for women not in union. We will use the package default data fpemdata::contraceptive_use for this example. Alternatively, you can supply a survey file path to fpemmodeling::do_1country_run and run your own data. The resulting samples can be combined to obtain results for all women. After obtaining all results we will plot some of the results.

options(warn=-1)

knitr::opts_chunk$set(echo = TRUE)
knitr::opts_chunk$set(fig.width=10, fig.height=7) 
  library(fpemdata)
  library(fpemdata)
  library(fpemmodeling)
  library(ggplot2)
  library(grid)
  library(gridExtra)

Run models

fpemmodeling::do_1country_run is a wrapper function to run the family planning estimation model for a coutnry of interest for either married or unmarried women. If the user provides data with missing values, the values will be imputed. If the user does not supply the first year arguement the function will choose a suitable year based on available data. The function takes in additional inputs such as services statistics if desired. See ??fpemmodeling::do_1country_run for all possible inputs. fpemmodeling::posterior_samples_all_women is used to combine samples to be used in the next step to obtain results for all women.

run_y <- fpemmodeling::do_1country_run(
  is_in_union = "Y",
  surveydata_filepath = NULL,
  service_stats = FALSE,
  division_numeric_code = 400,
  first_year = 1989,
  last_year = 2030
)
run_n <- fpemmodeling::do_1country_run(
  is_in_union = "N",
  surveydata_filepath = NULL,
  service_stats = FALSE,
  division_numeric_code = 400,
  first_year = 1989,
  last_year = 2030
)
core_data <- run_y$core_data
observations_y <- run_y$core_data$observations
observations_n <- run_n$core_data$observations
samples_all <- fpemmodeling::posterior_samples_all_women(in_union_posterior_samples = run_y$posterior_samples, 
                                                         not_in_union_posterior_samples = run_n$posterior_samples, 
                                                         core_data = core_data)

Process the samples

fpemreporting::fpem_calculate_results returns point-estimates for several indicators in long-format.

 population_counts <- fpemdata::population_counts %>%
      dplyr::filter(division_numeric_code == core_data$units$division_numeric_code)
  results_y <- fpemreporting::fpem_calculate_results(
    posterior_samples = run_y$posterior_samples,
    country_population_counts = population_counts %>%
      dplyr::filter(is_in_union == "Y"),
    first_year = min(core_data$time_frame$limits())
  )
  results_n <- fpemreporting::fpem_calculate_results(
    posterior_samples = run_n$posterior_samples,
    country_population_counts = population_counts %>%
      dplyr::filter(is_in_union == "N"),
    first_year = min(core_data$time_frame$limits())
  )
  results_all <- fpemreporting::fpem_calculate_results(
    posterior_samples = samples_all,
    country_population_counts = population_counts,
    first_year = min(core_data$time_frame$limits())
  )

Plot results for in-union, not-in-union, and all women

fpemreporting::fpem_plot_country_results plots the results of the model againts the observations. Choose any indicators returned from fpemreporting::fpem_calculate_results

indicators <- c(
    "unmet_need_any",
    "contraceptive_use_modern",
    "contraceptive_use_traditional",
    "contraceptive_use_any"
    )
plots <- fpemreporting::fpem_plot_country_results(
  country_results = results_y,
  observations = observations_y,
  first_year = core_data$time_frame$`.->.sequence` %>% min,
  last_year = core_data$time_frame$`.->.sequence` %>% max,
  is_in_union = "Y",
  indicators = indicators
  )
gridExtra::grid.arrange(grobs=plots[1:length(indicators)],
                 ncol=2,
                 top=textGrob("In-union women"))  

plots <- fpemreporting::fpem_plot_country_results(
  country_results = results_n,
  observations = observations_n,
  first_year = core_data$time_frame$`.->.sequence` %>% min,
  last_year = core_data$time_frame$`.->.sequence` %>% max,
  is_in_union = "N",
  indicators = indicators
  )
gridExtra::grid.arrange(grobs=plots[1:length(indicators)],
                 ncol=2,
                 top=textGrob("Not-in-union women"))  

plots <- fpemreporting::fpem_plot_country_results(
  country_results = results_all,
  observations = rbind(observations_n, observations_y),
  first_year = core_data$time_frame$`.->.sequence` %>% min,
  last_year = core_data$time_frame$`.->.sequence` %>% max,
  indicators = indicators
  )
gridExtra::grid.arrange(grobs=plots[1:length(indicators)],
                 ncol=2,
                 top=textGrob("All women"))       

FPRgroup/FPEM documentation built on March 3, 2020, 8:19 a.m.