Make regural expression classifications from LIKE% classifications

Source: R/make_regex.R

Make regural expression classifications from LIKE% classifications

make_regex(.data, classname, diagnosis, diagnosis.rm, sep = ", ")

Arguments

.data

holds the data which has variable names given as 'classname' and 'diagnosis'. Must be convertible to a tibble.
classname

name of the variable containing names/labels for the classes
diagnosis

name of the variable containing strings in LIKE% -format
diagnosis.rm

name of the variable defining exceptions of ICD-codes not to be included in the class
sep

separator of different diagnoses in 'diagnosis'. String.

Value

returns a tibble with regular expression definitions

See also

Other help functions: get_classification_name(), get_na_vars(), get_var_types(), left_join_replace_na(), replace_na_by_type()

Examples

if (FALSE) {
# We load classification here
excelfile <- readxl::read_excel("./datas/disease_classification.xlsx")
cl9 <- excelfile %>%
  make_regex(classname=Lyhenne,diagnosis=sairdiag_icd9,diagnosis.rm=sairdiagpl_icd9) %>%
  mutate(icd="icd9")
cl10 <- excelfile %>%
  make_regex(classname=Lyhenne,diagnosis=sairdiag,diagnosis.rm=sairdiagpl) %>%
  mutate(icd="icd10")

sel_classes<-bind_rows(cl9,cl10) %>%
  mutate(classification="own_definition")
}