DHSr (version 0.1.0)

Spdeplisa: Calculate Local Moran's I and Sign Combination Variables

Description

This function calculates Local Moran's I for a specified variable in a dataset and creates sign combination variables based on the standardized variable and the local Moran's I values.

Usage

Spdeplisa(data, variable_name, listw)

Value

A data frame containing the original data with additional columns:

lisa_I

Local Moran's I values for the specified variable.

lisa_p

P-values corresponding to the Local Moran's I values.

z_i

Standardized values of the input variable.

sign_combination2

Categories based on the sign of z_i and lisa_I (e.g., "positive-negative").

sign_combination3

Categories based on the sign of z_i and lisa_I (e.g., "High-High").

Arguments

data

A dataframe containing the spatial data.

variable_name

A string representing the name of the variable to be analyzed.

listw

A listw object containing spatial weights for the dataset.

Examples

# Load necessary libraries
if (requireNamespace("spData", quietly = TRUE)) {
  library(spData)
  library(sf)
  library(dplyr)

  # Use US states data as a substitute for a shapefile
  us_states <- spData::us_states

  # Simplify for demonstration: Select a subset of columns
  us_states_data <- us_states %>%
    select(GEOID, NAME) %>%
    mutate(mean_wealth = rnorm(nrow(us_states), 50, 10))  # Add mock data

  # Define a temporary shapefile path
  shapefile_path <- tempfile(fileext = ".shp")
  sf::st_write(us_states, shapefile_path, quiet = TRUE)

  # Create spatial weights using the listw function from the package
  us_states_listw <- DHSr::listw(
    shapefile_path = shapefile_path,
    data = us_states_data %>% sf::st_drop_geometry(),  # Drop geometry for compatibility
    loc_shape = "GEOID",
    loc_data = "GEOID",
    weight_function = function(d) exp(-d / 0.2)
  )

  # Apply the Spdeplisa function
  lisa_result <- DHSr::Spdeplisa(
    data = us_states_data,
    variable_name = "mean_wealth",
    listw = us_states_listw
  )

  # View the result
  head(lisa_result)
}