Skip to contents

Convert Config to Numeric Vector for Matrix Storage

Source: R/convert_config_to_matrix.R
convert_config_to_matrix.Rd

Converts a complex MOSAIC config list object directly into a numeric vector suitable for storage in a matrix. This function is optimized for calibration workflows where parameters need to be stored efficiently in pre-allocated matrices.

Usage

convert_config_to_matrix(config)

Arguments

config

A configuration list object in the format of MOSAIC::config_default. The config object should contain both global parameters (scalars) and location-specific parameters (vectors/lists).

Value

A named numeric vector where:

  • All parameter values are converted to numeric representation

  • Logical values become 0 (FALSE) or 1 (TRUE)

  • Character values are converted to NA with a warning

  • Vector parameters are expanded with appropriate suffixes

  • Names preserve the parameter structure for later reconstruction

Details

This function is more efficient than convert_config_to_dataframe() when the goal is to store parameters in a numeric matrix, as it avoids the overhead of creating an intermediate data.frame. It processes parameters directly into a numeric vector while maintaining the same naming convention.

The function handles:

  • Scalar parameters: Direct numeric conversion

  • Location-specific vectors: Expanded with location suffixes

  • Node-level parameters: Expanded with numeric indices

  • Logical values: Converted to 0/1

  • NULL values: Skipped

  • Non-numeric values: Converted to NA with warning

See also

convert_config_to_dataframe for data.frame output

Examples

if (FALSE) { # \dontrun{
# Convert sampled config for matrix storage
config_sampled <- sample_parameters(PATHS, seed = 123)
vec_params <- convert_config_to_matrix(config_sampled)

# Use in calibration workflow
n_sim <- 100
n_params <- length(vec_params)
results_matrix <- matrix(NA_real_, nrow = n_sim, ncol = 4 + n_params)
results_matrix[1, 5:ncol(results_matrix)] <- vec_params

# Compare with dataframe approach (slower); same columns, same order
df_params <- convert_config_to_dataframe(config_sampled)
all.equal(vec_params, unlist(df_params))  # TRUE
} # }