Skip to contents

Convert Wide Format to Long Format

Source: R/collapse.R
pivot_wide_to_long.Rd

Converts wide format data (where each antibiotic is a column) to long format (one row per organism-antibiotic combination). This is the first step before normalization and analysis.

Usage

pivot_wide_to_long(
  data,
  antibiotic_cols = NULL,
  pattern = NULL,
  id_cols = c("patient_id", "event_id", "organism_name", "date_of_culture"),
  antibiotic_name_col = "antibiotic_name",
  antibiotic_value_col = "antibiotic_value",
  remove_missing = TRUE,
  create_event_id = FALSE
)

Arguments

data

Data frame in wide format (antibiotics as columns)

antibiotic_cols

Character vector. Names of antibiotic columns to pivot. If NULL, will auto-detect based on pattern. Default NULL.

pattern

Character. Regex pattern to identify antibiotic columns if antibiotic_cols not provided. Default NULL (no auto-detection).

id_cols

Character vector. Columns to keep as identifiers (not pivoted). Default c("patient_id", "event_id", "organism_name", "date_of_culture").

antibiotic_name_col

Character. Name for the new column containing antibiotic names. Default "antibiotic_name".

antibiotic_value_col

Character. Name for the new column containing susceptibility results. Default "antibiotic_value".

remove_missing

Logical. Remove rows where antibiotic_value is NA, empty, or "-". Default TRUE.

create_event_id

Logical. Create event_id column if it doesn't exist (uses row numbers). Default FALSE.

Value

Data frame in long format

Examples

if (FALSE) { # \dontrun{
# Specify antibiotic columns explicitly
long_data <- pivot_wide_to_long(
  data = raw_data,
  antibiotic_cols = c("AMIKACIN", "GENTAMICIN", "CIPROFLOXACIN")
)

# Auto-detect columns by pattern (columns 12-53)
long_data <- pivot_wide_to_long(
  data = raw_data,
  antibiotic_cols = names(raw_data)[12:53]
)

# Auto-detect uppercase antibiotic names
long_data <- pivot_wide_to_long(
  data = raw_data,
  pattern = "^[A-Z]+$"
)
} # }