Skip to contents

Normalize indicator values

Source: R/normalization.R
normalize_indicators.Rd

Transforms indicator values to a common scale for comparison and aggregation.

Usage

normalize_indicators(
  data,
  indicators = NULL,
  method = c("minmax", "zscore", "quantile"),
  suffix = "_norm",
  keep_original = TRUE,
  na.rm = TRUE,
  reference_data = NULL,
  by_family = FALSE
)

Arguments

data

An sf object or data.frame containing indicator values

indicators

Character vector of indicator column names to normalize. If NULL, auto-detects indicator columns.

method

Character. Normalization method. Options:

  • "minmax" - Min-max normalization to 0-100 scale (default)

  • "zscore" - Z-score standardization (mean=0, sd=1)

  • "quantile" - Quantile normalization (0-100 based on percentile rank)

suffix

Character. Suffix to add to normalized column names. Default "_norm".

keep_original

Logical. Keep original indicator columns? Default TRUE.

na.rm

Logical. Remove NA values before normalization? Default TRUE.

reference_data

Optional data.frame with reference values for normalization. Useful for normalizing new data using parameters from a reference dataset.

by_family

Logical. If TRUE, normalize indicators within each family using family-wide parameters (e.g., all Carbon indicators C1, C2 share the same min/max). This makes indicators within a family directly comparable. Default FALSE.

Value

The input data with added normalized columns

Details

Normalization methods:

  • Min-max (0-100): norm = (value - min) / (max - min) * 100 - Preserves the original distribution shape - Sensitive to outliers - Interpretable scale (0 = worst, 100 = best)

  • Z-score: norm = (value - mean) / sd - Centers data around 0 - Units in standard deviations - Less sensitive to outliers

  • Quantile: norm = percentile_rank * 100 - Robust to outliers - Creates uniform distribution - 0 = lowest percentile, 100 = highest

See also

create_composite_index

Examples

if (FALSE) { # \dontrun{
# Normalize all indicators with min-max
normalized <- normalize_indicators(
  results,
  indicators = c("carbon", "biodiversity", "water"),
  method = "minmax"
)

# Z-score normalization
normalized_z <- normalize_indicators(
  results,
  method = "zscore",
  suffix = "_z"
)

# Normalize using reference dataset
new_normalized <- normalize_indicators(
  new_data,
  indicators = c("carbon", "water"),
  reference_data = reference_results
)
} # }