Draft R Code - achafetz/RetentionAnalysis GitHub Wiki

Psuedo coding for R analysis

District Names Crosswalk

  • import district crosswalk file

AIS data

  • import AIS data
  • merge AIS with district names (due to mismatch)
  • aggregate data to district level
  • recreate percent variable for aggregated districts

EA data

  • import PEPFAR FY16 expenditure data
  • keep only district level totals for retention expenditures
    • national_sub_unit datim_snu_id cbcts_lnkg_exp cbcts_rtnadhr_exp fbcts_loaded_tot
  • merge with district crosswalk file

MER data

  • import ICPI Fact View FY16 Q4 data
  • keep if operatingunit = Boswana and indicator = TX_RET (numerator & denominator)
  • aggregate to district level
  • reshape wide
  • create retention percentage by district

IMPATT data

  • import ICPI IMPATT data
  • keep if operatingunit = Boswana and indicator = PLHIV & TX_CURR_SUBNAT
  • reshape wide

Merge & Analyze

  • merge all datasets together
  • perform logit regression using survey, expenditure, and results data

Notes