idhs

Experimental R package for working with the IPUMS-processed DHS data (links and citations to come).

An example visualization here.

Install

# install.packages("remotes")
remotes::install_github("ki-tools/idhs")

Examples

library(idhs)

# preprocess an IPUMS extract
dd <- preprocess_ipums(
  input_xml = "ext-data/idhs/extract2/idhs_0000.xml",
  output_file = "ext-data/idhs/extract1/dd.rds"
)

# pull all shapefiles associated with data in the extract
preprocess_geo(dd, output_dir = "ext-data/idhs/geo/")

# html page showing variable names and their descriptions
view_var_descs(dd)
view_var_descs(dd, include_geo = TRUE)

# ---------- get regional yearly STI prevalence ---------- #

# numerator is stianyr (had STI in last 12 months) = "Yes"
attributes(dd$stianyr)[c("labels", "label")]
# denominator is all responses except sexactiv4wk = "never had intercourse"
attributes(dd$sexactiv4wk)[c("labels", "label")]

sti_rates <- calc_rates(dd,
  geo_dir = "ext-data/idhs/geo/",
  num_var = "stianyr",
  num_cond = 1,
  denom_var = "sexactiv4wk",
  denom_cond = c(1:9)
)

# -------------------- visualizations -------------------- #

# individual scatter
scatter_vis_all(
  sti_rates,
  title = "Women reporting an STI in the 12 months preceding the survey",
  subtitle = "Among women who ever had sexual intercourse, by country subregion",
  ylab = "Percentage of women reporting an STI"
)

# individual geo
geo_vis_all(sti_rates,
  title = "Percent STI for latest country survey",
  geo_dir = "ext-data/idhs/geo/",
  bins = c(0, 2, 4, 6, 10, 15, 35)
)

# joint
geo_scatter_vis(
  sti_rates,
  title = "Women reporting an STI in the 12 months preceding the survey",
  subtitle = "Among women who ever had sexual intercourse, by country subregion (hover a point to see percentage and 95% CI)",
  ylab = "Percentage of women reporting an STI",
  geo_title = "Percent STI for latest country survey",
  geo_dir = "ext-data/idhs/geo/",
  bins = c(0, 2, 4, 6, 10, 15, 35)
)