data_skim.R

## ###########################################################

## This script is copied from: 
## https://github.com/opensafely/comparative-booster/blob/fefa245aebba4cca792d991caebe0cb5d8d40de6/analysis/data_skim.R

## ** written by @wjchulme **
## ###########################################################

#################

# This script takes a dataset, summarises the variables using:
#   * skimr::skim() -- UNREDACTED,
#   * class(),
#   * and redacted tabulate functions
# and saves the output to a .txt file
# The script should only be run via an action in the project.yaml only
# The script must be accompanied by two arguments
# The first is the dataset, saved as an .rds file, that is to be summarised
# The second in the directory where the txt output will be saved

#################


# import libraries
library('tidyverse')
library('here')

source(here("analysis", "utils", "redaction.R")) # updated from original script due to no location

# import command-line arguments ----

args <- commandArgs(trailingOnly=TRUE)

if(length(args)==0){
  # use for interactive testing
  rds_file <- "output/processed/input_flowchart.rds"
  output_dir <- "output/data_properties"
} else {
  rds_file <- args[[1]]
  output_dir <- args[[2]]
}


stopifnot("must pass an .rds file" = fs::path_ext(rds_file)=="rds")

filenamebase <- fs::path_ext_remove(fs::path_file(rds_file))

# Import processed data ----

data <- readr::read_rds(here(rds_file))

# Output summary .txt ----

options(width=200) # set output width for capture.output

dir.create(here(output_dir), showWarnings = FALSE, recursive=TRUE)

## high-level variable overview ----
capture.output(
  skimr::skim_without_charts(data),
  file = here(output_dir, paste0(filenamebase, "_skim", ".txt")),
  split=FALSE
)

## list of column types ----
capture.output(
  lapply(data, class),
  file = here(output_dir, paste0(filenamebase, "_coltypes", ".txt"))
)


## tabulated data ----

# delete file if it exists
if(file.exists(here(output_dir, paste0(filenamebase, "_tabulate", ".txt")))){
  file.remove(here(output_dir, paste0(filenamebase, "_tabulate", ".txt")))
}


### categorical and logical ----
sumtabs_cat <-
  data %>%
  select(-ends_with("_id")) %>%
  select(where(is.character), where(is.logical), where(is.factor)) %>%
  map(redacted_summary_cat) %>%
  enframe()

capture.output(
  walk2(sumtabs_cat$value, sumtabs_cat$name, print_cat),
  file = here(output_dir, paste0(filenamebase, "_tabulate", ".txt")),
  append=FALSE
)


### numeric ----
sumtabs_num <-
  data %>%
  select(-ends_with("_id")) %>%
  select(where(~ {!is.logical(.x) & is.numeric(.x) & !is.Date(.x)})) %>%
  map(redacted_summary_num) %>%
  enframe()

capture.output(
  walk2(sumtabs_num$value, sumtabs_num$name, print_num),
  file = here(output_dir, paste0(filenamebase, "_tabulate", ".txt")),
  append=TRUE
)

### dates ----

sumtabs_date <-
  data %>%
  select(-ends_with("_id")) %>%
  select(where(is.Date)) %>%
  map(redacted_summary_date) %>%
  enframe()

capture.output(
  walk2(sumtabs_date$value, sumtabs_date$name, print_num),
  file = here(output_dir, paste0(filenamebase, "_tabulate", ".txt")),
  append=TRUE
)