When we received files with date variables, sometimes they are not in good 'date' format. Instead, they might be in pure numeric values with different digits. In such case, we have to check if they are ideal eight-digit number or not. Some digits may just not make sense and have to ask DBM about them. This function aims to make such inspection easier.
Arguments
- df_arg
Input data frame.
- id_string
A string contained in ID's column name.
- date_string
A string contained in date's column name.
- required_digits
The expected date's digit. Ideally it should be 8, i.e. yyyymmdd.
- required_leading_digit
The expected date's leading digit.
Value
A list with three elements:
- count_digit
A tibble of ID, date and date's digits.
- wrong_digit
A tibble of dates that do not have required digits.
- leading_digit
A tibble of dates with required digits and their leading number.
Details
Given a data frame, strings containing ID and date columns' name, and required digits, users can find the observations with wrong digits. For the dates with required digits, we can further check if they have required numbers, which in most cases should be "1" or "2" for eight-digit numbers.
The date_string argument should lead to a unique date column. This function only supports inspect one date column at one time.
Examples
if (FALSE) {
date_digit_info <- get_date_digit(
df_arg = fake_snack_df,
id_string = "Lop",
date_string = "numeric_date",
required_digits = 8, required_leading_digit = 2
)
# get the digits
date_digit_info$count_digit
# get the wrong dates
date_digit_info$wrong_digit
# check if the leading number is as required
date_digit_info$leading_digit
}