R/fun2_1_3_RegData_newVarMAsNA.R
In Rapporteket/norspis2: Plots and Tables of Results and Data Quality in NorSpis
Defines functions
fun2_1_3_RegData_newVarMAsNA
Documented in
fun2_1_3_RegData_newVarMAsNA
#' Make variables where missing is coded as NA (MAsNA)
#'
#' Output is data, with new variables where missing is coded as NA.
#' Some are categoricial, and som have values that are labels.
#' Accordingly the new variables are named with suffixes _MISSING, CAT, and _NAMES
#' (TODO: rename to for instance MaSNA, NumCat and LabCat)
#'
#'
#' Note: This function's name when writing aarsrapport 2019 (in 2020) was: make23_data_newVar
#'
#' @param myInData RegData
#'
#' @return myInData (invisble)
#' @export
#'
#' @examples
fun2_1_3_RegData_newVarMAsNA <- function(myInData){
#Made/formatted:
#Year
#PasientAlder_CAT_MISSING
#B01Sivilstatus_MISSING
#B01Sivilstatus_MISSING_NAMES
#B03Bosituasjon_MISSING
#B03Bosituasjon_MISSING_NAMES
#B02EgneBarn_MISSING
#B04PabegyntUtd_MISSING
#B05FullfortUtd_MISSING
#B06Hovedaktivitet_MISSING
#B07Hovedinntekt_MISSING
#B02EgneBarn_MISSING_NAMES
#B04PabegyntUtd_MISSING_NAMES
#B05FullfortUtd_MISSING_NAMES
#B06Hovedaktivitet_MISSING_NAMES
#B07Hovedinntekt_MISSING_NAMES
#MedBMI_CAT_MISSING
#DiagVSF_MISSING_NAMES
#DiagBUAkse1_MISSING_NAMES
#DiagSF_V_BU_MISSING_NAMES
#DiagDSM5Hoved_V_BU_MISSING_NAMES
#DiagVSomatiske_CAT_MISSING
#MedPsykofarmaka_CAT_MISSING
#MedAntidepressiva_CAT_MISSING
#MedNevroleptika_CAT_MISSING
#MedBenzodiazepiner_CAT_MISSING
#MedAnnenMedBeh_CAT_MISSING
#MedBlodprove_CAT_MISSING
#MedBeintetthMaling_CAT_MISSING
#B17FysMishandl_CAT_MISSING
#B18PsykMishandl_CAT_MISSING
#B19Overgrep_CAT_MISSING
#B20Mobbing_CAT_MISSING
#B21SelvskadTidl
#B22SelvskadSisteAr
#B23SelvmordFTidl
#B24SelvmordFSisteAr
#B25Avhengighet
#PT01OnsketInvolv
#pT02BleInvolv
#PT04KontaktBrukerorg
#PT05OrientertBrukerorg
#PO05Involvert_CAT_MISSING
#MedBMI_withIsoBMIBGSvalues
#MISSING(TODO:replace with "MAsNA??) means that we have specifed which values should be counted as missing if not present, and these values are given the value NA
#CAT means(TODO:replace with "NumCat"??)...a numeric variable that have been categorized.
#NAMES (TODO:replace with "LabCat" ??) means...it is with labeled categories
#"Year" variable
myInData$Year <- lubridate::year(myInData$HovedDato_FRMT)
#PasientAlder_CAT_MISSING (categorized and where missing is set to NA)
#this variable will by design have 100 % completeness
#(no "null" values and NAs must thus mean that registration is still in the making (BasisRegStatus is not equl to 1))
myInData$PasientAlder_CAT_MISSING <- cut(myInData$PasientAlder,
breaks=c(0,10,15,20,25,30,40,50, Inf),
labels=c("(0,10]","(10,15]","(15,20]","(20,25]","(25,30]","(30,40]","(40,50]","50+"),
ordered_result = T)
#B01 _MISSING
#new var where we decide and code which cases with value "null" should be interpreted as missing and thus coded into NA:
myInData$B01Sivilstatus_MISSING <- myInData$B01Sivilstatus#make new var similar to old
#myInData$B01Sivilstatus.xMISSING[myInData$B01Sivilstatus.x == 'null']<- NA #first code "null" to NA
myInData$B01Sivilstatus_MISSING[myInData$RegRegtype %in% c(1,2,3,4) & #then choose which that we should code as "missing"
#myInData$PasientAlder.x >= 15 &
myInData$B01Sivilstatus == 'null'] <- NA
#B01 _MISSING_NAMES
#coded the missing var into av var with text values
myInData$B01Sivilstatus_MISSING_NAMES <- dplyr::recode_factor(myInData$B01Sivilstatus_MISSING, #make it an ordered factor to prevent strange sorting
"1"="Enslig",
"2"="Samboer",
"3"="Gift",
"4"="Skilt",
"5"="Enke/enkemann",
"9"="Annen",
"null"="null",
.ordered = TRUE)
#B03
myInData$B03Bosituasjon_MISSING <- myInData$B03Bosituasjon
myInData$B03Bosituasjon_MISSING[myInData$RegRegtype %in% c(1,2,3,4) & #then choose which that we should code as "missing"
#myInData$PasientAlder.x >= 15 &
myInData$B01Sivilstatus == 'null'] <- NA
#B03 _MISSING_NAMES
myInData$B03Bosituasjon_MISSING_NAMES <- dplyr::recode_factor(myInData$B03Bosituasjon_MISSING,
"1"="Hos en av foreldrene",
"2"="Hos begge foreldrene",
"3"="Bor alene",
"4"="Med partner",
"5"="Med partner og barn",
"6"="Uten partner med barn",
"9"="Annen",
"null"="null",
.ordered = TRUE)
# B02,B04, B05, B06 and B07..._MISSING
varToRecode <- c("B02EgneBarn","B04PabegyntUtd","B05FullfortUtd","B06Hovedaktivitet", "B07Hovedinntekt")
for(varToRecode in varToRecode){
myInData[,paste0(varToRecode, "_MISSING")]<- myInData[, paste0(varToRecode)]#make new var similar to old
#myInData$B01Sivilstatus.xMISSING[myInData$B01Sivilstatus.x == 'null']<- NA #first code "null" to NA
myInData[,paste0(varToRecode, "_MISSING")][myInData$RegRegtype %in% c(1,2,3,4) & #CONDITIONS TO CODE AS MISSING (startreg and age>15, and where value is "null")
myInData$PasientAlder >= 15 &
myInData[,varToRecode] == 'null'] <- NA
}#OUTPUT: This gives out the variables with _MISSING suffix.
# B02 _MISSING_NAMES (NEW CATEGORIES)
myInData$B02EgneBarn_MISSING_NAMES <- replace(myInData$B02EgneBarn_MISSING,
myInData$B02EgneBarn_MISSING >= "1" &
!(myInData$B02EgneBarn_MISSING =="null") #values above 1 and not "null" categorized as "1+"
,
"1+")
#make it an ordered factor as other _MISSING_NAMES variables:
myInData$B02EgneBarn_MISSING_NAMES <- factor(myInData$B02EgneBarn_MISSING_NAMES,
levels = c("0","1+", "null"),
ordered = TRUE)
# B04 _MISSING_NAMES
myInData$B04PabegyntUtd_MISSING_NAMES <- dplyr::recode_factor(myInData$B04PabegyntUtd_MISSING,
"1"="Grunnskole",
"2"="Videreg?ende skole (1-3 ?r)",
"3"="H?gskole eller universitet, mindre enn 4 ?r",
"4"="H?gskole eller universitet, 4 ?r eller mer",
"9"="Ukjent",
"null"="null",
.ordered = TRUE)
# B05 _MISSING_NAMES
myInData$B05FullfortUtd_MISSING_NAMES <- dplyr::recode_factor(myInData$B05FullfortUtd_MISSING,
"1"="Ikke fullf?rt grunnskole",
"2"="Grunnskole",
"3"="Videreg?ende skole (1-3 ?r)",
"4"="H?gskole eller universitet, mindre enn 4 ?r",
"5"="H?gskole eller universitet, 4 ?r eller mer",
"9"="Ukjent",
"null"="null",
.ordered = TRUE)
# B06 _MISSING_NAMES
myInData$B06Hovedaktivitet_MISSING_NAMES <- dplyr::recode_factor(myInData$B06Hovedaktivitet_MISSING,
"1"="Heltidsarbeid",
"2"="Deltidsarbeid",
"3"="P? arbeidsmarkedstiltak",
"4"="Vernepliktig",
"5"="Skoleelev/l?rling",
"6"="Student",
"7"="Sykemeldt",
"8"="Uf?r",
"9"="Annen",
"null"="null",
.ordered = TRUE)
# B07 _MISSING_NAMES
myInData$B07Hovedinntekt_MISSING_NAMES <- dplyr::recode_factor(myInData$B07Hovedinntekt_MISSING,
"1"="Arbeidsinntekt",
"2"="Sykepenger/trygd/pensjon",
"3"="Blir fors?rget",
"4"="Sosialhjelp",
"5"="Stipend/l?n",
"6"="Kursst?nad/l?nn i arbeidsmarkedstiltak",
"9"="Andre inntekter",
"null"="null",
.ordered = TRUE)
# MEDICAL CHARACTERISTICS
#MedBMI_CAT_MISSING (categorized and where missing is set to NA)
myInData$MedBMI_CAT_MISSING <- cut(myInData$MedBMI, #this variable will be design have 100 % completeness (no "null" values and NAs must be interpreted as e.g. not yet completely registered registration)
breaks=c(0,10,12.5,15,17.5,20,25,30, Inf),
labels=c("(0,10]","(10,12.5]","(12.5,15]","(15,17.5]","(17.5,20]","(20,25]","(25,30]","30+"),
ordered_result = T)
# DiagVSF_MISSING_NAMES (with missing coded as NA and categories as NAMES)
myInData <- myInData %>% dplyr::mutate(DiagVSF_MISSING_NAMES = dplyr::case_when(
DiagVSF == "F500" ~ "F50.0",
DiagVSF == "F501" ~ "F50.1",
DiagVSF == "F502" ~ "F50.2",
DiagVSF == "F503" ~ "F50.3",
DiagVSF == "F504" ~ "F50.4",
DiagVSF == "F505" ~ "F50.5",
DiagVSF == "F508" ~ "F50.8",
DiagVSF == "F509" ~ "F50.9",
DiagVSF == "F982" ~ "F98.2",
DiagVSF == "F983" ~ "F98.3",
RegRegtype %in% c(1,5,98) & DiagVSF == "" ~ NA_character_,
DiagVSF == "" ~ "null",
TRUE ~ "Annen"))#this line code all other diagnoses as "annen"
myInData$DiagVSF_MISSING_NAMES <- factor(myInData$DiagVSF_MISSING_NAMES, #make it an ordered factor
levels = c("F50.0",
"F50.1",
"F50.2",
"F50.3",
"F50.4",
"F50.5",
"F50.8",
"F50.9",
"F98.2",
"F98.3",
"Annen",
"null"),
ordered = TRUE)
#DiagBUAkse1_MISSING_NAMES
#Code values containing the different diagnoses
myInData <- myInData %>% dplyr::mutate(DiagBUAkse1_MISSING_NAMES = dplyr::case_when(
grepl("F500", DiagBUAkse1, fixed = T) == T ~ "F50.0",
grepl("F501", DiagBUAkse1, fixed = T) == T ~ "F50.1",
grepl("F502", DiagBUAkse1, fixed = T) == T ~ "F50.2",
grepl("F503", DiagBUAkse1, fixed = T) == T ~ "F50.3",
grepl("F504", DiagBUAkse1, fixed = T) == T ~ "F50.4",
grepl("F505", DiagBUAkse1, fixed = T) == T ~ "F50.5",
grepl("F508", DiagBUAkse1, fixed = T) == T ~ "F50.8",
grepl("F509", DiagBUAkse1, fixed = T) == T ~ "F50.9",
grepl("F982", DiagBUAkse1, fixed = T) == T ~ "F98.2",
grepl("F983", DiagBUAkse1, fixed = T) == T ~ "F98.3",
(DiagBUAkse1 == "" | DiagBUAkse1 == 1999) & RegRegtype %in% c(2,6,99) ~ NA_character_,
DiagBUAkse1 == "" ~ "null",
TRUE ~"Annen"))
#Gj?r om tll factor:
myInData$DiagBUAkse1_MISSING_NAMES <- factor(myInData$DiagBUAkse1_MISSING_NAMES, #make it an ordered factor
levels = c("F50.0",
"F50.1",
"F50.2",
"F50.3",
"F50.4",
"F50.5",
"F50.8",
"F50.9",
"F98.2",
"F98.3",
"Annen",
"null"),
ordered = TRUE)
# DiagSF_V_BU_MISSING_NAMES (with missing coded as NA and categories as NAMES) (DiagVSF and DiagBUAkse1 combined)
myInData <- myInData %>% dplyr::mutate(DiagSF_V_BU_MISSING_NAMES = dplyr::case_when(
DiagVSF_MISSING_NAMES == "F50.0"| DiagBUAkse1_MISSING_NAMES == "F50.0" ~ "F50.0" ,
DiagVSF_MISSING_NAMES == "F50.1"| DiagBUAkse1_MISSING_NAMES == "F50.1" ~ "F50.1",
DiagVSF_MISSING_NAMES == "F50.2"| DiagBUAkse1_MISSING_NAMES == "F50.2" ~ "F50.2",
DiagVSF_MISSING_NAMES == "F50.3"| DiagBUAkse1_MISSING_NAMES == "F50.3" ~ "F50.3",
DiagVSF_MISSING_NAMES == "F50.4"| DiagBUAkse1_MISSING_NAMES == "F50.4" ~ "F50.4",
DiagVSF_MISSING_NAMES == "F50.5"| DiagBUAkse1_MISSING_NAMES == "F50.5" ~ "F50.5",
DiagVSF_MISSING_NAMES == "F50.8"| DiagBUAkse1_MISSING_NAMES == "F50.8" ~ "F50.8",
DiagVSF_MISSING_NAMES == "F50.9"| DiagBUAkse1_MISSING_NAMES == "F50.9" ~ "F50.9",
DiagVSF_MISSING_NAMES == "F98.2"| DiagBUAkse1_MISSING_NAMES == "F98.2" ~ "F98.2",
DiagVSF_MISSING_NAMES == "F98.3"| DiagBUAkse1_MISSING_NAMES == "F98.3" ~ "F98.3",
RegRegtype %in% c(1,2,5,6,98,99) & DiagVSF == "" & (DiagBUAkse1 == "" | DiagBUAkse1 == 1999) ~ NA_character_,
DiagVSF == "" & (DiagBUAkse1 == "" | DiagBUAkse1 == 1999) ~ "null",
TRUE ~ "Annen"))#this line code all other diagnoses as "annen"
myInData$DiagSF_V_BU_MISSING_NAMES <- factor(myInData$DiagSF_V_BU_MISSING_NAMES, #make it an ordered factor
levels = c("F50.0",
"F50.1",
"F50.2",
"F50.3",
"F50.4",
"F50.5",
"F50.8",
"F50.9",
"F98.2",
"F98.3",
"Annen",
"null"),
ordered = TRUE)
#DiagDSM5Hoved_V_BU_MISSING_NAMES (DiagVDSM5Hoved and DiagBUDSM5Hoved combined )
myInData <- myInData %>% dplyr::mutate(DiagDSM5Hoved_V_BU_MISSING_NAMES = dplyr::case_when(
DiagVDSM5Hoved == "0" | DiagBUDSM5Hoved == "0" ~ "307.52 Pica (ICD-10: F50.8)", #"307.52 \n Pica \n (ICD-10: F50.8)", #"307.52 \n Pica \n (ICD-10: F50.8)",
DiagVDSM5Hoved == "1" | DiagBUDSM5Hoved == "1" ~ "307.53 Dr?vtygging (ICD-10: F98.21)", #"307.53 \n Dr?vtygging \n (ICD-10: F98.21)", #"307.53 \n Dr?vtygging \n (ICD-10: F98.21)",
DiagVDSM5Hoved == "2" | DiagBUDSM5Hoved == "2" ~ "307.59 Unnv./restr. (ICD-10: F50.8)",#"307.59 \n Unnv./restr. \n (ICD-10: F50.8)",#"307.59 \n Unnv./restr. \n (ICD-10: F50.8)",
DiagVDSM5Hoved == "3" | DiagBUDSM5Hoved == "3" ~ "307.1 Anorexia N. (ICD-10: F50.0)", #"307.1 \n Anorexia N. \n (ICD-10: F50.0)", #"307.1 \n Anorexia N. \n (ICD-10: F50.0)",
DiagVDSM5Hoved == "4" | DiagBUDSM5Hoved == "4" ~ "307.51 Bulimia N. (ICD-10: F50.2)", #"307.51 \n Bulimia N. \n (ICD-10: F50.2)", #"307.51 \n Bulimia N. \n (ICD-10: F50.2)",
DiagVDSM5Hoved == "5" | DiagBUDSM5Hoved == "5" ~ "307.51 Overspisingsl. (ICD-10: F50.8)",#"307.51 \n Overspisingsl. \n (ICD-10: F50.8)",#"307.51 \n Overspisingsl. \n (ICD-10: F50.8)",#klokebok sier "overspisningsforstyrrelse", nasjonale retn.linjer *lidelse, derfor forkortet til "*l."
DiagVDSM5Hoved == "6" | DiagBUDSM5Hoved == "6" ~ "307.59 Annen spesifisert (ICD-10: F50.8)", #"307.59 \n Annen spesifisert \n (ICD-10: F50.8)", #"307.59 \n Annen spesifisert \n (ICD-10: F50.8)",
DiagVDSM5Hoved == "7" | DiagBUDSM5Hoved == "7" ~ "307.50 Uspesifisert (ICD-10: F50.9)",#"307.50 \n Uspesifisert \n (ICD-10: F50.9)",#"307.50 \n Uspesifisert \n (ICD-10: F50.9)",
RegRegtype %in% c(1,2,5,6,98,99) & DiagVDSM5Hoved == "null" & DiagBUDSM5Hoved == "null" ~ NA_character_, #NA valueas are give those observations that should have had values
DiagVDSM5Hoved == "null" & DiagBUDSM5Hoved == "null" ~ "null",
TRUE ~ "Annen"))#this line code all other diagnoses as "annen"
myInData$DiagDSM5Hoved_V_BU_MISSING_NAMES <- factor(myInData$DiagDSM5Hoved_V_BU_MISSING_NAMES, #make it an ordered factor
levels = c("307.52 Pica (ICD-10: F50.8)", #"307.52 \n Pica \n (ICD-10: F50.8)",
"307.53 Dr?vtygging (ICD-10: F98.21)", #"307.53 \n Dr?vtygging \n (ICD-10: F98.21)",
"307.59 Unnv./restr. (ICD-10: F50.8)",#"307.59 \n Unnv./restr. \n (ICD-10: F50.8)",
"307.1 Anorexia N. (ICD-10: F50.0)", #"307.1 \n Anorexia N. \n (ICD-10: F50.0)",
"307.51 Bulimia N. (ICD-10: F50.2)", #"307.51 \n Bulimia N. \n (ICD-10: F50.2)",
"307.51 Overspisingsl. (ICD-10: F50.8)",#"307.51 \n Overspisingsl. \n (ICD-10: F50.8)",
"307.59 Annen spesifisert (ICD-10: F50.8)", #"307.59 \n Annen spesifisert \n (ICD-10: F50.8)",
"307.50 Uspesifisert (ICD-10: F50.9)",#"307.50 \n Uspesifisert \n (ICD-10: F50.9)",
"Annen", #"Annen",
"null"),#"null"),
ordered = TRUE)
#DiagVSomatiske_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(DiagVSomatiske_CAT_MISSING = dplyr::case_when(
DiagVSomatiske == "0" ~ "Ingen relevant somatiske",
DiagVSomatiske == "1" & DiagVMalabsorpsjon == "1" ~ "Malabsorpsj.tilstander",
DiagVSomatiske == "1" & DiagVDiabetes == "1" ~ "Diabetes",
DiagVSomatiske == "1" & DiagVDiabetes == "0" & DiagVMalabsorpsjon == "0" ~ "Ja, men ikke spesifisert",
DiagVSomatiske == "null" ~ "null",
RegRegtype %in% c(1,5,98) & DiagVSomatiske == "null" ~ NA_character_)
)
#Gj?r om tll factor:
myInData$DiagVSomatiske_CAT_MISSING <- factor(myInData$DiagVSomatiske_CAT_MISSING, #make it an ordered factor
levels = c("Diabetes",
"Malabsorpsj.tilstander",
"Ja, men ikke spesifisert",
"Ingen relevant somatiske",
"null"="null"),
ordered = TRUE)
#MedPsykofarmaka_CAT_MISSING Psykofarmakologisk behandling
myInData <- myInData %>% dplyr::mutate(MedPsykofarmaka_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" ~ "Nei",
MedPsykofarmaka == "1" ~ "Ja",
MedPsykofarmaka == "null" ~ "null"))
#MedAntidepressiva_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedAntidepressiva_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedAntidepressiva == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAntidepressiva == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAntidepressiva == "1" ~ "Ja",
MedAntidepressiva == "null" ~ "null"))
#MedNevroleptika_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedNevroleptika_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedNevroleptika == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedNevroleptika == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedNevroleptika == "1" ~ "Ja",
MedNevroleptika == "null" ~ "null"))
#MedBenzodiazepiner_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedBenzodiazepiner_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedBenzodiazepiner == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedBenzodiazepiner == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedBenzodiazepiner == "1" ~ "Ja",
MedBenzodiazepiner == "null" ~ "null"))
#MedAnnenMedBeh_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedAnnenMedBeh_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedAnnenMedBeh == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAnnenMedBeh == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAnnenMedBeh == "1" ~ "Ja",
MedAnnenMedBeh == "null" ~ "null"))
#MedBlodprove_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedBlodprove_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedBlodprove == "null" ~ NA_character_,
MedBlodprove == "0" ~ "Nei",
MedBlodprove == "1" ~ "Ja",
MedBlodprove == "9" ~ "Ikke aktuelt",
MedBlodprove == "null" ~ "null"))
#ordered factor
myInData$MedBlodprove_CAT_MISSING <- factor(myInData$MedBlodprove_CAT_MISSING, #make it an ordered factor
levels = c("Nei",
"Ja",
"Ikke aktuelt",
"null"),
ordered = TRUE)
#MedBeintetthMaling_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedBeintetthMaling_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedBeintetthMaling == "null" ~ NA_character_,
MedBeintetthMaling == "0" ~ "Nei",
MedBeintetthMaling == "1" ~ "Ja",
MedBeintetthMaling == "9" ~ "Vet ikke",
MedBeintetthMaling == "null" ~ "null"))
#ordered factor
myInData$MedBeintetthMaling_CAT_MISSING <- factor(myInData$MedBeintetthMaling_CAT_MISSING, #make it an ordered factor
levels = c("Nei",
"Ja",
"Vet ikke",
"null"),
ordered = TRUE)
#B17FysMishandl
myInData <- myInData %>% dplyr::mutate(B17FysMishandl_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B17FysMishandl == "0" ~ "Nei",
B17FysMishandl == "1" ~ "Ja",
B17FysMishandl == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B17FysMishandl == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B17FysMishandl == "null" ~"null"))
#B18PsykMishandl
myInData <- myInData %>% dplyr::mutate(B18PsykMishandl_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B18PsykMishandl == "0" ~ "Nei",
B18PsykMishandl == "1" ~ "Ja",
B18PsykMishandl == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B18PsykMishandl == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B18PsykMishandl == "null" ~"null"))
#B19Overgrep
myInData <- myInData %>% dplyr::mutate(B19Overgrep_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B19Overgrep == "0" ~ "Nei",
B19Overgrep == "1" ~ "Ja",
B19Overgrep == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B19Overgrep == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B19Overgrep == "null" ~"null"))
#B20Mobbing
myInData <- myInData %>% dplyr::mutate(B20Mobbing_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B20Mobbing == "0" ~ "Nei",
B20Mobbing == "1" ~ "Ja",
B20Mobbing == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B20Mobbing == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B20Mobbing == "null" ~ "null"))
#B21SelvskadTidl,B22SelvskadSisteAr, B23SelvmordFTidl
##doing as above, but more effectively, using a for loop
##FIX/IMPROVE: currently the suffix _CAT_MISSING is not added, so original variable is overwritten (we want to keep original var.)
myVarQuo <- c(rlang::quo(B21SelvskadTidl),
rlang::quo(B22SelvskadSisteAr),
rlang::quo(B23SelvmordFTidl),
rlang::quo(B24SelvmordFSisteAr),
rlang::quo(B25Avhengighet))
for(i in 1:length(myVarQuo)){
myInData <- myInData %>% dplyr::mutate(!!myVarQuo[[i]] := dplyr::case_when(#paste0(!!myVarQuo[[i]],"_CAT_MISSING") := case_when
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
!!myVarQuo[[i]] == "0" ~ "Nei",
!!myVarQuo[[i]] == "1" ~ "Ja",
!!myVarQuo[[i]] == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
!!myVarQuo[[i]] == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
!!myVarQuo[[i]] == "null" ~ "null"))
}
#PT01OnsketInvolv,
#pT02BleInvolv,
#PT04KontaktBrukerorg,
#PT05OrientertBrukerorg,
##FIX/IMPROVE: currently the suffix _CAT_MISSING is not added, so original variable is overwritten (we want to keep original var.)
##...so that it becomes:
#PT01OnsketInvolv_01_MISSING,
#PT02BleInvolv_01_MISSING,
#PT04KontaktBrukerorg_01_MISSING,
#PT05OrientertBrukerorg_01_MISSING,
myVarQuo2 <- c(rlang::quo(PT01OnsketInvolv),
rlang::quo(PT02BleInvolv),
rlang::quo(PT04KontaktBrukerorg),
rlang::quo(PT05OrientertBrukerorg))
for(i in 1:length(myVarQuo2)){
myInData <- myInData %>% dplyr::mutate(!!myVarQuo2[[i]] := dplyr::case_when(
RegRegtype %in% c(5,6) & PasientAlder >= 16 & !!myVarQuo2[[i]] == "null" ~ NA_character_,
!!myVarQuo2[[i]] == "0" ~ "0",
!!myVarQuo2[[i]] == "1" ~ "1",
!!myVarQuo2[[i]] == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
!!myVarQuo2[[i]] == "null" ~ "null"))
}
##Pasienterfaringer (PasOpp)
#PO05Involvert_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(PO05Involvert_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
PO05Involvert == "0" ~ "Ikke i det hele tatt",
PO05Involvert == "1" ~ "I liten grad",
PO05Involvert == "2" ~ "I noen grad",
PO05Involvert == "3" ~ "I stor grad",
PO05Involvert == "4" ~ "I sv?rt stor grad",
PO05Involvert == "9" ~ "Ikke aktuelt",
PO05Involvert == "null" & RegRegtype %in% c(5,6) ~ NA_character_,
PO05Involvert == "null" ~ "null"))
#ordered factor
myInData$PO05Involvert_CAT_MISSING <- factor(myInData$PO05Involvert_CAT_MISSING, #make it an ordered factor
levels = c("Ikke i det hele tatt",
"I liten grad",
"I noen grad",
"I stor grad",
"I sv?rt stor grad"),
#"Ikke aktuelt"),
ordered = TRUE)
#MedBMI_withIsoBMIBGSvalues (MedBMI but with MedISOBMIBGS values for those that have ISOBMIBGS values)
myInData <- myInData %>% dplyr::mutate(MedBMI_withIsoBMIBGSvalues = dplyr::case_when(
!is.na(MedIsoBMIBGS) ~ MedIsoBMIBGS, #those with ISOBMIBGS values get ISoBMIBGS values
TRUE ~ MedBMI)) #rest gets ordinary BMI values
#TODO remember a quality control of this variable, as we had to change
#the code to take into account NA (versus "null" before) when going
#into production with MAR.
output <- myInData
return(invisible(output))
}
Rapporteket/norspis2 documentation built on April 18, 2021, 1:11 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions fun2_1_3_RegData_newVarMAsNA
Documented in fun2_1_3_RegData_newVarMAsNA
#' Make variables where missing is coded as NA (MAsNA)
#'
#' Output is data, with new variables where missing is coded as NA.
#' Some are categoricial, and som have values that are labels.
#' Accordingly the new variables are named with suffixes _MISSING, CAT, and _NAMES
#' (TODO: rename to for instance MaSNA, NumCat and LabCat)
#'
#'
#' Note: This function's name when writing aarsrapport 2019 (in 2020) was: make23_data_newVar
#'
#' @param myInData RegData
#'
#' @return myInData (invisble)
#' @export
#'
#' @examples
fun2_1_3_RegData_newVarMAsNA <- function(myInData){
#Made/formatted:
#Year
#PasientAlder_CAT_MISSING
#B01Sivilstatus_MISSING
#B01Sivilstatus_MISSING_NAMES
#B03Bosituasjon_MISSING
#B03Bosituasjon_MISSING_NAMES
#B02EgneBarn_MISSING
#B04PabegyntUtd_MISSING
#B05FullfortUtd_MISSING
#B06Hovedaktivitet_MISSING
#B07Hovedinntekt_MISSING
#B02EgneBarn_MISSING_NAMES
#B04PabegyntUtd_MISSING_NAMES
#B05FullfortUtd_MISSING_NAMES
#B06Hovedaktivitet_MISSING_NAMES
#B07Hovedinntekt_MISSING_NAMES
#MedBMI_CAT_MISSING
#DiagVSF_MISSING_NAMES
#DiagBUAkse1_MISSING_NAMES
#DiagSF_V_BU_MISSING_NAMES
#DiagDSM5Hoved_V_BU_MISSING_NAMES
#DiagVSomatiske_CAT_MISSING
#MedPsykofarmaka_CAT_MISSING
#MedAntidepressiva_CAT_MISSING
#MedNevroleptika_CAT_MISSING
#MedBenzodiazepiner_CAT_MISSING
#MedAnnenMedBeh_CAT_MISSING
#MedBlodprove_CAT_MISSING
#MedBeintetthMaling_CAT_MISSING
#B17FysMishandl_CAT_MISSING
#B18PsykMishandl_CAT_MISSING
#B19Overgrep_CAT_MISSING
#B20Mobbing_CAT_MISSING
#B21SelvskadTidl
#B22SelvskadSisteAr
#B23SelvmordFTidl
#B24SelvmordFSisteAr
#B25Avhengighet
#PT01OnsketInvolv
#pT02BleInvolv
#PT04KontaktBrukerorg
#PT05OrientertBrukerorg
#PO05Involvert_CAT_MISSING
#MedBMI_withIsoBMIBGSvalues
#MISSING(TODO:replace with "MAsNA??) means that we have specifed which values should be counted as missing if not present, and these values are given the value NA
#CAT means(TODO:replace with "NumCat"??)...a numeric variable that have been categorized.
#NAMES (TODO:replace with "LabCat" ??) means...it is with labeled categories
#"Year" variable
myInData$Year <- lubridate::year(myInData$HovedDato_FRMT)
#PasientAlder_CAT_MISSING (categorized and where missing is set to NA)
#this variable will by design have 100 % completeness
#(no "null" values and NAs must thus mean that registration is still in the making (BasisRegStatus is not equl to 1))
myInData$PasientAlder_CAT_MISSING <- cut(myInData$PasientAlder,
breaks=c(0,10,15,20,25,30,40,50, Inf),
labels=c("(0,10]","(10,15]","(15,20]","(20,25]","(25,30]","(30,40]","(40,50]","50+"),
ordered_result = T)
#B01 _MISSING
#new var where we decide and code which cases with value "null" should be interpreted as missing and thus coded into NA:
myInData$B01Sivilstatus_MISSING <- myInData$B01Sivilstatus#make new var similar to old
#myInData$B01Sivilstatus.xMISSING[myInData$B01Sivilstatus.x == 'null']<- NA #first code "null" to NA
myInData$B01Sivilstatus_MISSING[myInData$RegRegtype %in% c(1,2,3,4) & #then choose which that we should code as "missing"
#myInData$PasientAlder.x >= 15 &
myInData$B01Sivilstatus == 'null'] <- NA
#B01 _MISSING_NAMES
#coded the missing var into av var with text values
myInData$B01Sivilstatus_MISSING_NAMES <- dplyr::recode_factor(myInData$B01Sivilstatus_MISSING, #make it an ordered factor to prevent strange sorting
"1"="Enslig",
"2"="Samboer",
"3"="Gift",
"4"="Skilt",
"5"="Enke/enkemann",
"9"="Annen",
"null"="null",
.ordered = TRUE)
#B03
myInData$B03Bosituasjon_MISSING <- myInData$B03Bosituasjon
myInData$B03Bosituasjon_MISSING[myInData$RegRegtype %in% c(1,2,3,4) & #then choose which that we should code as "missing"
#myInData$PasientAlder.x >= 15 &
myInData$B01Sivilstatus == 'null'] <- NA
#B03 _MISSING_NAMES
myInData$B03Bosituasjon_MISSING_NAMES <- dplyr::recode_factor(myInData$B03Bosituasjon_MISSING,
"1"="Hos en av foreldrene",
"2"="Hos begge foreldrene",
"3"="Bor alene",
"4"="Med partner",
"5"="Med partner og barn",
"6"="Uten partner med barn",
"9"="Annen",
"null"="null",
.ordered = TRUE)
# B02,B04, B05, B06 and B07..._MISSING
varToRecode <- c("B02EgneBarn","B04PabegyntUtd","B05FullfortUtd","B06Hovedaktivitet", "B07Hovedinntekt")
for(varToRecode in varToRecode){
myInData[,paste0(varToRecode, "_MISSING")]<- myInData[, paste0(varToRecode)]#make new var similar to old
#myInData$B01Sivilstatus.xMISSING[myInData$B01Sivilstatus.x == 'null']<- NA #first code "null" to NA
myInData[,paste0(varToRecode, "_MISSING")][myInData$RegRegtype %in% c(1,2,3,4) & #CONDITIONS TO CODE AS MISSING (startreg and age>15, and where value is "null")
myInData$PasientAlder >= 15 &
myInData[,varToRecode] == 'null'] <- NA
}#OUTPUT: This gives out the variables with _MISSING suffix.
# B02 _MISSING_NAMES (NEW CATEGORIES)
myInData$B02EgneBarn_MISSING_NAMES <- replace(myInData$B02EgneBarn_MISSING,
myInData$B02EgneBarn_MISSING >= "1" &
!(myInData$B02EgneBarn_MISSING =="null") #values above 1 and not "null" categorized as "1+"
,
"1+")
#make it an ordered factor as other _MISSING_NAMES variables:
myInData$B02EgneBarn_MISSING_NAMES <- factor(myInData$B02EgneBarn_MISSING_NAMES,
levels = c("0","1+", "null"),
ordered = TRUE)
# B04 _MISSING_NAMES
myInData$B04PabegyntUtd_MISSING_NAMES <- dplyr::recode_factor(myInData$B04PabegyntUtd_MISSING,
"1"="Grunnskole",
"2"="Videreg?ende skole (1-3 ?r)",
"3"="H?gskole eller universitet, mindre enn 4 ?r",
"4"="H?gskole eller universitet, 4 ?r eller mer",
"9"="Ukjent",
"null"="null",
.ordered = TRUE)
# B05 _MISSING_NAMES
myInData$B05FullfortUtd_MISSING_NAMES <- dplyr::recode_factor(myInData$B05FullfortUtd_MISSING,
"1"="Ikke fullf?rt grunnskole",
"2"="Grunnskole",
"3"="Videreg?ende skole (1-3 ?r)",
"4"="H?gskole eller universitet, mindre enn 4 ?r",
"5"="H?gskole eller universitet, 4 ?r eller mer",
"9"="Ukjent",
"null"="null",
.ordered = TRUE)
# B06 _MISSING_NAMES
myInData$B06Hovedaktivitet_MISSING_NAMES <- dplyr::recode_factor(myInData$B06Hovedaktivitet_MISSING,
"1"="Heltidsarbeid",
"2"="Deltidsarbeid",
"3"="P? arbeidsmarkedstiltak",
"4"="Vernepliktig",
"5"="Skoleelev/l?rling",
"6"="Student",
"7"="Sykemeldt",
"8"="Uf?r",
"9"="Annen",
"null"="null",
.ordered = TRUE)
# B07 _MISSING_NAMES
myInData$B07Hovedinntekt_MISSING_NAMES <- dplyr::recode_factor(myInData$B07Hovedinntekt_MISSING,
"1"="Arbeidsinntekt",
"2"="Sykepenger/trygd/pensjon",
"3"="Blir fors?rget",
"4"="Sosialhjelp",
"5"="Stipend/l?n",
"6"="Kursst?nad/l?nn i arbeidsmarkedstiltak",
"9"="Andre inntekter",
"null"="null",
.ordered = TRUE)
# MEDICAL CHARACTERISTICS
#MedBMI_CAT_MISSING (categorized and where missing is set to NA)
myInData$MedBMI_CAT_MISSING <- cut(myInData$MedBMI, #this variable will be design have 100 % completeness (no "null" values and NAs must be interpreted as e.g. not yet completely registered registration)
breaks=c(0,10,12.5,15,17.5,20,25,30, Inf),
labels=c("(0,10]","(10,12.5]","(12.5,15]","(15,17.5]","(17.5,20]","(20,25]","(25,30]","30+"),
ordered_result = T)
# DiagVSF_MISSING_NAMES (with missing coded as NA and categories as NAMES)
myInData <- myInData %>% dplyr::mutate(DiagVSF_MISSING_NAMES = dplyr::case_when(
DiagVSF == "F500" ~ "F50.0",
DiagVSF == "F501" ~ "F50.1",
DiagVSF == "F502" ~ "F50.2",
DiagVSF == "F503" ~ "F50.3",
DiagVSF == "F504" ~ "F50.4",
DiagVSF == "F505" ~ "F50.5",
DiagVSF == "F508" ~ "F50.8",
DiagVSF == "F509" ~ "F50.9",
DiagVSF == "F982" ~ "F98.2",
DiagVSF == "F983" ~ "F98.3",
RegRegtype %in% c(1,5,98) & DiagVSF == "" ~ NA_character_,
DiagVSF == "" ~ "null",
TRUE ~ "Annen"))#this line code all other diagnoses as "annen"
myInData$DiagVSF_MISSING_NAMES <- factor(myInData$DiagVSF_MISSING_NAMES, #make it an ordered factor
levels = c("F50.0",
"F50.1",
"F50.2",
"F50.3",
"F50.4",
"F50.5",
"F50.8",
"F50.9",
"F98.2",
"F98.3",
"Annen",
"null"),
ordered = TRUE)
#DiagBUAkse1_MISSING_NAMES
#Code values containing the different diagnoses
myInData <- myInData %>% dplyr::mutate(DiagBUAkse1_MISSING_NAMES = dplyr::case_when(
grepl("F500", DiagBUAkse1, fixed = T) == T ~ "F50.0",
grepl("F501", DiagBUAkse1, fixed = T) == T ~ "F50.1",
grepl("F502", DiagBUAkse1, fixed = T) == T ~ "F50.2",
grepl("F503", DiagBUAkse1, fixed = T) == T ~ "F50.3",
grepl("F504", DiagBUAkse1, fixed = T) == T ~ "F50.4",
grepl("F505", DiagBUAkse1, fixed = T) == T ~ "F50.5",
grepl("F508", DiagBUAkse1, fixed = T) == T ~ "F50.8",
grepl("F509", DiagBUAkse1, fixed = T) == T ~ "F50.9",
grepl("F982", DiagBUAkse1, fixed = T) == T ~ "F98.2",
grepl("F983", DiagBUAkse1, fixed = T) == T ~ "F98.3",
(DiagBUAkse1 == "" | DiagBUAkse1 == 1999) & RegRegtype %in% c(2,6,99) ~ NA_character_,
DiagBUAkse1 == "" ~ "null",
TRUE ~"Annen"))
#Gj?r om tll factor:
myInData$DiagBUAkse1_MISSING_NAMES <- factor(myInData$DiagBUAkse1_MISSING_NAMES, #make it an ordered factor
levels = c("F50.0",
"F50.1",
"F50.2",
"F50.3",
"F50.4",
"F50.5",
"F50.8",
"F50.9",
"F98.2",
"F98.3",
"Annen",
"null"),
ordered = TRUE)
# DiagSF_V_BU_MISSING_NAMES (with missing coded as NA and categories as NAMES) (DiagVSF and DiagBUAkse1 combined)
myInData <- myInData %>% dplyr::mutate(DiagSF_V_BU_MISSING_NAMES = dplyr::case_when(
DiagVSF_MISSING_NAMES == "F50.0"| DiagBUAkse1_MISSING_NAMES == "F50.0" ~ "F50.0" ,
DiagVSF_MISSING_NAMES == "F50.1"| DiagBUAkse1_MISSING_NAMES == "F50.1" ~ "F50.1",
DiagVSF_MISSING_NAMES == "F50.2"| DiagBUAkse1_MISSING_NAMES == "F50.2" ~ "F50.2",
DiagVSF_MISSING_NAMES == "F50.3"| DiagBUAkse1_MISSING_NAMES == "F50.3" ~ "F50.3",
DiagVSF_MISSING_NAMES == "F50.4"| DiagBUAkse1_MISSING_NAMES == "F50.4" ~ "F50.4",
DiagVSF_MISSING_NAMES == "F50.5"| DiagBUAkse1_MISSING_NAMES == "F50.5" ~ "F50.5",
DiagVSF_MISSING_NAMES == "F50.8"| DiagBUAkse1_MISSING_NAMES == "F50.8" ~ "F50.8",
DiagVSF_MISSING_NAMES == "F50.9"| DiagBUAkse1_MISSING_NAMES == "F50.9" ~ "F50.9",
DiagVSF_MISSING_NAMES == "F98.2"| DiagBUAkse1_MISSING_NAMES == "F98.2" ~ "F98.2",
DiagVSF_MISSING_NAMES == "F98.3"| DiagBUAkse1_MISSING_NAMES == "F98.3" ~ "F98.3",
RegRegtype %in% c(1,2,5,6,98,99) & DiagVSF == "" & (DiagBUAkse1 == "" | DiagBUAkse1 == 1999) ~ NA_character_,
DiagVSF == "" & (DiagBUAkse1 == "" | DiagBUAkse1 == 1999) ~ "null",
TRUE ~ "Annen"))#this line code all other diagnoses as "annen"
myInData$DiagSF_V_BU_MISSING_NAMES <- factor(myInData$DiagSF_V_BU_MISSING_NAMES, #make it an ordered factor
levels = c("F50.0",
"F50.1",
"F50.2",
"F50.3",
"F50.4",
"F50.5",
"F50.8",
"F50.9",
"F98.2",
"F98.3",
"Annen",
"null"),
ordered = TRUE)
#DiagDSM5Hoved_V_BU_MISSING_NAMES (DiagVDSM5Hoved and DiagBUDSM5Hoved combined )
myInData <- myInData %>% dplyr::mutate(DiagDSM5Hoved_V_BU_MISSING_NAMES = dplyr::case_when(
DiagVDSM5Hoved == "0" | DiagBUDSM5Hoved == "0" ~ "307.52 Pica (ICD-10: F50.8)", #"307.52 \n Pica \n (ICD-10: F50.8)", #"307.52 \n Pica \n (ICD-10: F50.8)",
DiagVDSM5Hoved == "1" | DiagBUDSM5Hoved == "1" ~ "307.53 Dr?vtygging (ICD-10: F98.21)", #"307.53 \n Dr?vtygging \n (ICD-10: F98.21)", #"307.53 \n Dr?vtygging \n (ICD-10: F98.21)",
DiagVDSM5Hoved == "2" | DiagBUDSM5Hoved == "2" ~ "307.59 Unnv./restr. (ICD-10: F50.8)",#"307.59 \n Unnv./restr. \n (ICD-10: F50.8)",#"307.59 \n Unnv./restr. \n (ICD-10: F50.8)",
DiagVDSM5Hoved == "3" | DiagBUDSM5Hoved == "3" ~ "307.1 Anorexia N. (ICD-10: F50.0)", #"307.1 \n Anorexia N. \n (ICD-10: F50.0)", #"307.1 \n Anorexia N. \n (ICD-10: F50.0)",
DiagVDSM5Hoved == "4" | DiagBUDSM5Hoved == "4" ~ "307.51 Bulimia N. (ICD-10: F50.2)", #"307.51 \n Bulimia N. \n (ICD-10: F50.2)", #"307.51 \n Bulimia N. \n (ICD-10: F50.2)",
DiagVDSM5Hoved == "5" | DiagBUDSM5Hoved == "5" ~ "307.51 Overspisingsl. (ICD-10: F50.8)",#"307.51 \n Overspisingsl. \n (ICD-10: F50.8)",#"307.51 \n Overspisingsl. \n (ICD-10: F50.8)",#klokebok sier "overspisningsforstyrrelse", nasjonale retn.linjer *lidelse, derfor forkortet til "*l."
DiagVDSM5Hoved == "6" | DiagBUDSM5Hoved == "6" ~ "307.59 Annen spesifisert (ICD-10: F50.8)", #"307.59 \n Annen spesifisert \n (ICD-10: F50.8)", #"307.59 \n Annen spesifisert \n (ICD-10: F50.8)",
DiagVDSM5Hoved == "7" | DiagBUDSM5Hoved == "7" ~ "307.50 Uspesifisert (ICD-10: F50.9)",#"307.50 \n Uspesifisert \n (ICD-10: F50.9)",#"307.50 \n Uspesifisert \n (ICD-10: F50.9)",
RegRegtype %in% c(1,2,5,6,98,99) & DiagVDSM5Hoved == "null" & DiagBUDSM5Hoved == "null" ~ NA_character_, #NA valueas are give those observations that should have had values
DiagVDSM5Hoved == "null" & DiagBUDSM5Hoved == "null" ~ "null",
TRUE ~ "Annen"))#this line code all other diagnoses as "annen"
myInData$DiagDSM5Hoved_V_BU_MISSING_NAMES <- factor(myInData$DiagDSM5Hoved_V_BU_MISSING_NAMES, #make it an ordered factor
levels = c("307.52 Pica (ICD-10: F50.8)", #"307.52 \n Pica \n (ICD-10: F50.8)",
"307.53 Dr?vtygging (ICD-10: F98.21)", #"307.53 \n Dr?vtygging \n (ICD-10: F98.21)",
"307.59 Unnv./restr. (ICD-10: F50.8)",#"307.59 \n Unnv./restr. \n (ICD-10: F50.8)",
"307.1 Anorexia N. (ICD-10: F50.0)", #"307.1 \n Anorexia N. \n (ICD-10: F50.0)",
"307.51 Bulimia N. (ICD-10: F50.2)", #"307.51 \n Bulimia N. \n (ICD-10: F50.2)",
"307.51 Overspisingsl. (ICD-10: F50.8)",#"307.51 \n Overspisingsl. \n (ICD-10: F50.8)",
"307.59 Annen spesifisert (ICD-10: F50.8)", #"307.59 \n Annen spesifisert \n (ICD-10: F50.8)",
"307.50 Uspesifisert (ICD-10: F50.9)",#"307.50 \n Uspesifisert \n (ICD-10: F50.9)",
"Annen", #"Annen",
"null"),#"null"),
ordered = TRUE)
#DiagVSomatiske_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(DiagVSomatiske_CAT_MISSING = dplyr::case_when(
DiagVSomatiske == "0" ~ "Ingen relevant somatiske",
DiagVSomatiske == "1" & DiagVMalabsorpsjon == "1" ~ "Malabsorpsj.tilstander",
DiagVSomatiske == "1" & DiagVDiabetes == "1" ~ "Diabetes",
DiagVSomatiske == "1" & DiagVDiabetes == "0" & DiagVMalabsorpsjon == "0" ~ "Ja, men ikke spesifisert",
DiagVSomatiske == "null" ~ "null",
RegRegtype %in% c(1,5,98) & DiagVSomatiske == "null" ~ NA_character_)
)
#Gj?r om tll factor:
myInData$DiagVSomatiske_CAT_MISSING <- factor(myInData$DiagVSomatiske_CAT_MISSING, #make it an ordered factor
levels = c("Diabetes",
"Malabsorpsj.tilstander",
"Ja, men ikke spesifisert",
"Ingen relevant somatiske",
"null"="null"),
ordered = TRUE)
#MedPsykofarmaka_CAT_MISSING Psykofarmakologisk behandling
myInData <- myInData %>% dplyr::mutate(MedPsykofarmaka_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" ~ "Nei",
MedPsykofarmaka == "1" ~ "Ja",
MedPsykofarmaka == "null" ~ "null"))
#MedAntidepressiva_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedAntidepressiva_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedAntidepressiva == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAntidepressiva == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAntidepressiva == "1" ~ "Ja",
MedAntidepressiva == "null" ~ "null"))
#MedNevroleptika_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedNevroleptika_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedNevroleptika == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedNevroleptika == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedNevroleptika == "1" ~ "Ja",
MedNevroleptika == "null" ~ "null"))
#MedBenzodiazepiner_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedBenzodiazepiner_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedBenzodiazepiner == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedBenzodiazepiner == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedBenzodiazepiner == "1" ~ "Ja",
MedBenzodiazepiner == "null" ~ "null"))
#MedAnnenMedBeh_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedAnnenMedBeh_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedPsykofarmaka == "null" ~ NA_character_,
MedPsykofarmaka == "0" & MedAnnenMedBeh == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAnnenMedBeh == "0" ~ "Nei",
MedPsykofarmaka == "1" & MedAnnenMedBeh == "1" ~ "Ja",
MedAnnenMedBeh == "null" ~ "null"))
#MedBlodprove_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedBlodprove_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedBlodprove == "null" ~ NA_character_,
MedBlodprove == "0" ~ "Nei",
MedBlodprove == "1" ~ "Ja",
MedBlodprove == "9" ~ "Ikke aktuelt",
MedBlodprove == "null" ~ "null"))
#ordered factor
myInData$MedBlodprove_CAT_MISSING <- factor(myInData$MedBlodprove_CAT_MISSING, #make it an ordered factor
levels = c("Nei",
"Ja",
"Ikke aktuelt",
"null"),
ordered = TRUE)
#MedBeintetthMaling_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(MedBeintetthMaling_CAT_MISSING = dplyr::case_when(
RegRegtype %in% c(1,2,3,4,5,6) & MedBeintetthMaling == "null" ~ NA_character_,
MedBeintetthMaling == "0" ~ "Nei",
MedBeintetthMaling == "1" ~ "Ja",
MedBeintetthMaling == "9" ~ "Vet ikke",
MedBeintetthMaling == "null" ~ "null"))
#ordered factor
myInData$MedBeintetthMaling_CAT_MISSING <- factor(myInData$MedBeintetthMaling_CAT_MISSING, #make it an ordered factor
levels = c("Nei",
"Ja",
"Vet ikke",
"null"),
ordered = TRUE)
#B17FysMishandl
myInData <- myInData %>% dplyr::mutate(B17FysMishandl_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B17FysMishandl == "0" ~ "Nei",
B17FysMishandl == "1" ~ "Ja",
B17FysMishandl == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B17FysMishandl == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B17FysMishandl == "null" ~"null"))
#B18PsykMishandl
myInData <- myInData %>% dplyr::mutate(B18PsykMishandl_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B18PsykMishandl == "0" ~ "Nei",
B18PsykMishandl == "1" ~ "Ja",
B18PsykMishandl == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B18PsykMishandl == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B18PsykMishandl == "null" ~"null"))
#B19Overgrep
myInData <- myInData %>% dplyr::mutate(B19Overgrep_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B19Overgrep == "0" ~ "Nei",
B19Overgrep == "1" ~ "Ja",
B19Overgrep == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B19Overgrep == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B19Overgrep == "null" ~"null"))
#B20Mobbing
myInData <- myInData %>% dplyr::mutate(B20Mobbing_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
B20Mobbing == "0" ~ "Nei",
B20Mobbing == "1" ~ "Ja",
B20Mobbing == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
B20Mobbing == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
B20Mobbing == "null" ~ "null"))
#B21SelvskadTidl,B22SelvskadSisteAr, B23SelvmordFTidl
##doing as above, but more effectively, using a for loop
##FIX/IMPROVE: currently the suffix _CAT_MISSING is not added, so original variable is overwritten (we want to keep original var.)
myVarQuo <- c(rlang::quo(B21SelvskadTidl),
rlang::quo(B22SelvskadSisteAr),
rlang::quo(B23SelvmordFTidl),
rlang::quo(B24SelvmordFSisteAr),
rlang::quo(B25Avhengighet))
for(i in 1:length(myVarQuo)){
myInData <- myInData %>% dplyr::mutate(!!myVarQuo[[i]] := dplyr::case_when(#paste0(!!myVarQuo[[i]],"_CAT_MISSING") := case_when
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
!!myVarQuo[[i]] == "0" ~ "Nei",
!!myVarQuo[[i]] == "1" ~ "Ja",
!!myVarQuo[[i]] == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
!!myVarQuo[[i]] == "null" & RegRegtype %in% c(1,2,3,4) ~ NA_character_,
!!myVarQuo[[i]] == "null" ~ "null"))
}
#PT01OnsketInvolv,
#pT02BleInvolv,
#PT04KontaktBrukerorg,
#PT05OrientertBrukerorg,
##FIX/IMPROVE: currently the suffix _CAT_MISSING is not added, so original variable is overwritten (we want to keep original var.)
##...so that it becomes:
#PT01OnsketInvolv_01_MISSING,
#PT02BleInvolv_01_MISSING,
#PT04KontaktBrukerorg_01_MISSING,
#PT05OrientertBrukerorg_01_MISSING,
myVarQuo2 <- c(rlang::quo(PT01OnsketInvolv),
rlang::quo(PT02BleInvolv),
rlang::quo(PT04KontaktBrukerorg),
rlang::quo(PT05OrientertBrukerorg))
for(i in 1:length(myVarQuo2)){
myInData <- myInData %>% dplyr::mutate(!!myVarQuo2[[i]] := dplyr::case_when(
RegRegtype %in% c(5,6) & PasientAlder >= 16 & !!myVarQuo2[[i]] == "null" ~ NA_character_,
!!myVarQuo2[[i]] == "0" ~ "0",
!!myVarQuo2[[i]] == "1" ~ "1",
!!myVarQuo2[[i]] == "9" ~ NA_character_, #9 is NA because it is the value "registrar" choose when patient has not answered
!!myVarQuo2[[i]] == "null" ~ "null"))
}
##Pasienterfaringer (PasOpp)
#PO05Involvert_CAT_MISSING
myInData <- myInData %>% dplyr::mutate(PO05Involvert_CAT_MISSING = dplyr::case_when(
#RegRegtype %in% c(1,2,3,4) & B17FysMishandl == "null" ~ NA_character_, #should not be necessary as 9 is missing value for this question
PO05Involvert == "0" ~ "Ikke i det hele tatt",
PO05Involvert == "1" ~ "I liten grad",
PO05Involvert == "2" ~ "I noen grad",
PO05Involvert == "3" ~ "I stor grad",
PO05Involvert == "4" ~ "I sv?rt stor grad",
PO05Involvert == "9" ~ "Ikke aktuelt",
PO05Involvert == "null" & RegRegtype %in% c(5,6) ~ NA_character_,
PO05Involvert == "null" ~ "null"))
#ordered factor
myInData$PO05Involvert_CAT_MISSING <- factor(myInData$PO05Involvert_CAT_MISSING, #make it an ordered factor
levels = c("Ikke i det hele tatt",
"I liten grad",
"I noen grad",
"I stor grad",
"I sv?rt stor grad"),
#"Ikke aktuelt"),
ordered = TRUE)
#MedBMI_withIsoBMIBGSvalues (MedBMI but with MedISOBMIBGS values for those that have ISOBMIBGS values)
myInData <- myInData %>% dplyr::mutate(MedBMI_withIsoBMIBGSvalues = dplyr::case_when(
!is.na(MedIsoBMIBGS) ~ MedIsoBMIBGS, #those with ISOBMIBGS values get ISoBMIBGS values
TRUE ~ MedBMI)) #rest gets ordinary BMI values
#TODO remember a quality control of this variable, as we had to change
#the code to take into account NA (versus "null" before) when going
#into production with MAR.
output <- myInData
return(invisible(output))
}
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.