9 Asset benefit indicator (ABI)
In this section, we harmonize variables needed to compute asset benefit indicator.
library(haven)
library(labelled) # for general functions to work with labelled data
library(tidyverse) # general wrangling
library(dplyr)
library(Hmisc)
library(gtsummary) # to demonstrate automatic use of variable labels in summary tables
library(readxl)
library(foreign)
library(sjPlot)
library(sjmisc)
library(sjlabelled) # for example efc data set with variable labels
library(stringr)dir_input_data = "C:/Users/AHema/OneDrive - CGIAR/Desktop/WFP Resilience dataset/data/input_data/Chad"
dir_output_data = "C:/Users/AHema/OneDrive - CGIAR/Desktop/WFP Resilience dataset/data/output_data/Chad"Chad_Harmonization_variables <- read_excel(paste0(dir_input_data,"/Chad_Harmonization.xlsx"),
sheet = "variables_harmonization")
#View(Chad_Harmonization_variables)
Chad_Harmonization_description <- read_excel(paste0(dir_input_data,"/Chad_Harmonization.xlsx"),
sheet = "description")
#View(Chad_Harmonization_description)lst_data = Chad_Harmonization_description$Data
lst_test = Chad_Harmonization_description$Name
for(i in 1:length(lst_data)) { # Head of for-loop
assign(lst_test[i], # Read and store data frames
read_sav(paste0(dir_input_data,"/",lst_data[i])))
}for (j in 1:length(lst_test)){
df= get(lst_test[j], envir = .GlobalEnv)
for (i in 1:nrow(Chad_Harmonization_variables)){
df[,Chad_Harmonization_variables$NewVariable_Name[i]]=ifelse(is.na(Chad_Harmonization_variables[i,lst_test[j]]),NA,df[,Chad_Harmonization_variables[i,lst_test[j]][[1]]])
}
df<-df %>% select(Chad_Harmonization_variables$NewVariable_Name)
assign(lst_test[j], # Read and store data frames
df)
#write_sav(df, paste0(dir_output_data,"/",lst_test[j],".sav"))
#write_dta(df, paste0(dir_output_data,"/",lst_test[j],".dta"))
}abi_variables = Chad_baseline_2018 %>% dplyr::select(gtsummary::starts_with("ABI")) %>% names()
abi_variables <- c(abi_variables,"ActifCreationEmploi",
"BeneficieEmploi",
"TRavailMaintienActif")
#
abi_variables <- abi_variables[! abi_variables %in% c('ABISexparticipant')]9.1 EA 2019
df_2019 = Chad_ea_2019 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_2019 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_2019$ABIProteger)
# Chad_ea_2019 <- Chad_ea_2019 %>%
# dplyr::mutate(across(abi_variables,
# ~recode(.,
# "Non" = 0,
# "Oui" = 1,
# "NA" = 888
# )
# )
# )
df_2019 <- df_2019 %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ as.character(.))) %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ dplyr::recode(., "Non" = 0, "Oui" = 1,"NA" = 888)))
df_2019 <- df_2019 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_2019$ABIProteger)
# haven::write_dta(df_2019,paste0(dir_output_data,"/","df_2019.dta"))9.2 EA 2020
df_2020 = Chad_ea_2020 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_2020 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_2020$ABIProteger)
df_2020 <- df_2020 %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ as.character(.))) %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ dplyr::recode(., "0" = 0, "1" = 1))) %>%
tidyr::replace_na(setNames(as.list(rep(888, length(abi_variables_tmp))), abi_variables_tmp))
df_2020 <- df_2020 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_2020$ABIProteger)
# haven::write_dta(df_2020,paste0(dir_output_data,"/","df_2020.dta"))9.3 EA 2021
df_2021 = Chad_ea_2021 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_2021 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_2021$ABIProteger)
df_2021 <- df_2021 %>%
dplyr::mutate(across(abi_variables_tmp,
~recode(.,
"1" = 0,
"2" = 1,
"3" = 888
)
)
)
df_2021 <- df_2021 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_2021$ABIProteger)9.4 EA 2022
df_2022 = Chad_ea_2022 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_2022 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_2022$ABIProteger)
df_2022 <- df_2022 %>%
dplyr::mutate(across(abi_variables_tmp,
~recode(.,
"1" = 0,
"2" = 1,
"3" = 888
)
)
)
df_2022 <- df_2022 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_2022$ABIProteger)9.5 EA 2023
df_2023 = Chad_ea_2023 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_2023 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_2023$ABIProteger)
# df_2023 <- df_2023 %>%
# dplyr::mutate(across(abi_variables_tmp,
# ~recode(.,
# "1" = 0,
# "2" = 1,
# "3" = 888
# )
# )
# )
df_2023 <- df_2023 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_2023$ABIProteger)9.6 PDM 2020
df_pdm_2020 = Chad_pdm_2020 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_pdm_2020 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_pdm_2020$ABIProteger)
df_pdm_2020 <- df_pdm_2020 %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ as.character(.))) %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ dplyr::recode(., "0" = 0, "1" = 1)))
df_pdm_2020 <- df_pdm_2020 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_pdm_2020$ABIProteger)9.7 PDM 2021
df_pdm_2021 = Chad_pdm_2021 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_pdm_2021 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_pdm_2021$ABIProteger)
df_pdm_2021 <- df_pdm_2021 %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ as.character(.))) %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ dplyr::recode(., "0" = 0, "1" = 1, "2" = 888)))
df_pdm_2021 <- df_pdm_2021 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_pdm_2021$ABIProteger)9.8 PDM 2022
df_pdm_2022 = Chad_pdm_2022 %>% labelled::to_factor() %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_pdm_2022 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
expss::val_lab(df_pdm_2022$ABIProteger)
df_pdm_2022 <- df_pdm_2022 %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ as.character(.))) %>%
dplyr::mutate(across(all_of(abi_variables_tmp), ~ dplyr::recode(., "Non" = 0, "Oui" = 1, "Ne sait pas" = 888)))
df_pdm_2022 <- df_pdm_2022 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_pdm_2022$ABIProteger)9.9 PDM 2023
df_pdm_2023 = Chad_pdm_2023 %>%
dplyr::select(ID,abi_variables)
all_missing_cols <- df_pdm_2023 %>%
select(where(~ all(is.na(.)))) %>% names()
#all_missing_cols
abi_variables_tmp <- abi_variables[! abi_variables %in% all_missing_cols]
df_pdm_2023 <- df_pdm_2023 %>%
dplyr::mutate(across(abi_variables_tmp,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_pdm_2023$ABIProteger)9.10 Merging all data
df_Chad<-plyr::rbind.fill(df_2019,
df_2020,
df_2021,
df_2022,
df_2023,
df_pdm_2020,
df_pdm_2021,
df_pdm_2022,
df_pdm_2023)expss::val_lab(df_Chad$ABIProteger)
not_ok = c("ABIParticipation","ABItransferts","ABITensions","ActifCreationEmploi","BeneficieEmploi","TRavailMaintienActif")
df_Chad <- df_Chad %>%
dplyr::mutate(across(not_ok,
~labelled(., labels = c(
"Non" = 0,
"Oui" = 1,
"Ne sait pas" = 888
)
)
)
)
expss::val_lab(df_Chad$ABIProteger)df_Chad$HHFFAPart <- df_Chad$ABIParticipation
df_Chad$HHAssetProtect <- df_Chad$ABIProteger
df_Chad$HHAssetProduct <- df_Chad$ABIProduction
df_Chad$HHAssetDecHardship <- df_Chad$ABIdifficultes
df_Chad$HHAssetAccess <- df_Chad$ABIMarches
df_Chad$HHTrainingAsset <- df_Chad$ABIGererActifs
df_Chad$HHAssetEnv <- df_Chad$ABIEnvironnement
df_Chad$HHWorkAsset <- df_Chad$ABIutiliseractifsdf_Chad <- df_Chad %>%
mutate(across(c(HHAssetProtect,HHAssetProduct,HHAssetDecHardship,HHAssetAccess,HHTrainingAsset,HHAssetEnv,HHWorkAsset), ~labelled(., labels = c(
"No" = 0,
"Yes" = 1,
"Don't know" = 888,
"Not applicable" = 999
))))
#recode 888/999 to 0
df_Chad <- df_Chad %>%
mutate(across(HHAssetProtect:HHWorkAsset, ~ dplyr::recode(.x, "0" = 0, "1" = 1, "999" = 0, "888" = 0)))
# Count non-missing values per row
df_Chad$NonMissingCount <- rowSums(!is.na(df_Chad[, c("ABIProteger", "ABIProduction",
"ABIdifficultes", "ABIMarches",
"ABIGererActifs", "ABIEnvironnement",
"ABIutiliseractifs")]))
#sum ABI rows
df_Chad <- df_Chad %>%
mutate(ABIScore = ifelse(NonMissingCount >0 ,rowSums(across(c(HHAssetProtect:HHWorkAsset)),na.rm = T),NA))#assign variable and value labels
var_label(df_Chad$HHFFAPart) <- "Have you or any of your household member participated in the asset creation activities and received a food assistance transfer?"
var_label(df_Chad$HHAssetProtect) <- "Do you think that the assets that were built or rehabilitated in your community are better protecting your household, its belongings and its production capacities (fields, equipment, etc.) from floods / drought / landslides / mudslides?"
var_label(df_Chad$HHAssetProduct) <- "Do you think that the assets that were built or rehabilitated in your community have allowed your household to increase or diversify its production (agriculture / livestock / other)?"
var_label(df_Chad$HHAssetDecHardship) <- "Do you think that the assets that were built or rehabilitated in your community have decreased the day-to-day hardship and released time for any of your family members (including women and children)?"
var_label(df_Chad$HHAssetAccess) <- "Do you think that the assets that were built or rehabilitated in your community have improved the ability of any of your household member to access markets and/or basic services (water, sanitation, health, education, etc)?"
var_label(df_Chad$HHTrainingAsset) <- "Do you think that the trainings and other support provided in your community have improved your household’s ability to manage and maintain assets?"
var_label(df_Chad$HHAssetEnv) <- "Do you think that the assets that were built or rehabilitated in your community have improved your natural environment (for example more vegetal cover, water table increased, less erosion, etc.)?"
var_label(df_Chad$HHWorkAsset) <- "Do you think that the works undertaken in your community have restored your ability to access and/or use basic asset functionalities?"