Install from GitHub with:
# install.packages("remotes")
remotes::install_github("epicentre-msf/datadict")The dict_from_odk() function can be used to generate an OCA-style data
dictionary from an ODK template (both the ‘survey’ and ‘options’ sheets
of the ODK template are required as inputs).
library(datadict)
library(readxl)
# path to example ODK template (a WHO mortality survey)
path_data <- system.file("extdata", package = "datadict")
path_odk_template <- file.path(path_data, "WHOVA2016_v1_5_3_ODK.xlsx")
# read 'survey' sheet and 'choices' sheet
odk_survey <- readxl::read_xlsx(path_odk_template, sheet = "survey")
odk_choices <- readxl::read_xlsx(path_odk_template, sheet = "choices")
# derive OCA-style data dictionary
dict <- dict_from_odk(odk_survey, odk_choices)
# examine first few rows/cols
dict[1:5,1:4]
#> # A tibble: 5 × 4
#> variable_name short_label type choices
#> <chr> <chr> <chr> <chr>
#> 1 Id10002 Is this a region of high HIV/AIDS mortality? Coded list high, High | low, Low | veryl, Very low
#> 2 Id10003 Is this a region of high malaria mortality? Coded list high, High | low, Low | veryl, Very low
#> 3 Id10004 During which season did (s)he die? Coded list wet, Wet | dry, Dry | DK, Doesn't know
#> 4 Id10007 What is the name of VA respondent? Free text <NA>
#> 5 Id10007a What is the sex of VA respondent? Coded list female, Female | male, Male | undetermined, Amb…The dict_from_redcap() function can be used to generate an OCA-style
data dictionary from a REDCap data dictionary. The input dictionary can
be exported directly from a REDCap project website or fetched via the
API using e.g. the R package
redcap.
# path to example REDCap template
path_data <- system.file("extdata", package = "datadict")
path_redcap_dict <- file.path(path_data, "REDCapDataDictionaryDemo.csv")
# read dictionary
redcap_dict <- read.csv(path_redcap_dict)
# derive OCA-style data dictionary
dict <- dict_from_redcap(redcap_dict)
# examine first few rows/cols
dict[1:5,1:5]
#> # A tibble: 5 × 5
#> variable_name short_label type choices origin
#> <chr> <chr> <chr> <chr> <chr>
#> 1 study_id Study ID Free text <NA> original
#> 2 date_enrolled Date subject signed consent Date <NA> original
#> 3 first_name First Name Free text <NA> original
#> 4 last_name Last Name Free text <NA> original
#> 5 address Street, City, State, ZIP Free text <NA> originalThe dict_from_data() function can be used to generate a template
OCA-style data dictionary (which may require further processing) from a
dataset. Data types are based on the class of each column within in the
input dataset, e.g.:
| Column class in R | Dictionary data type |
|---|---|
| Date | Date |
| POSIX | Datetime |
| logical | Logical |
| integer | Numeric |
| numeric | Numeric |
| factor | Coded list |
| character | Coded list or Free text (see argument factor_threshold) |
# path to example dataset
path_data <- system.file("extdata", package = "datadict")
path_linelist <- file.path(path_data, "linelist_cleaned.xlsx")
# read data
dat <- readxl::read_xlsx(path_linelist)
# derive OCA-style data dictionary template
dict <- dict_from_data(dat)
# examine first few rows/cols
dict[1:7,1:5]
#> # A tibble: 7 × 5
#> variable_name short_label type choices origin
#> <chr> <chr> <chr> <chr> <chr>
#> 1 case_id <NA> Free text <NA> original
#> 2 generation <NA> Numeric <NA> original
#> 3 cohort_fu <NA> Logical <NA> original
#> 4 date_infection <NA> Datetime <NA> original
#> 5 date_onset <NA> Datetime <NA> original
#> 6 date_hospitalisation <NA> Datetime <NA> original
#> 7 date_outcome <NA> Datetime <NA> original