Linear Access to Neuroimaging Data (Methods for neuroim2 Generic)

These methods provide 4D linear access to data for specific `fmristore` classes, implementing the linear_access generic function from the neuroim2 package.

The `linear_access` generic allows direct access to data elements using a single numeric index that spans the entire 4D space of the object (X, Y, Z, Time). Refer to the documentation for linear_access in the neuroim2 package for general details about the generic concept.

Provides 4D linear access to data in an H5ParcellatedScan object. It reconstructs voxel values on the fly from the HDF5 file based on their cluster assignments. Values for voxels outside the mask are returned as 0.

Usage

linear_access(x, i, ...)

# S4 method for class 'H5ParcellatedScan,numeric'
linear_access(x, i, ...)

# S4 method for class 'H5ParcellatedScanSummary,numeric'
linear_access(x, i, ...)

# S4 method for class 'H5NeuroVec,numeric'
linear_access(x, i)

# S4 method for class 'H5NeuroVec,integer'
linear_access(x, i)

# S4 method for class 'H5NeuroVol,numeric'
linear_access(x, i)

# S4 method for class 'H5NeuroVol,integer'
linear_access(x, i)

# S4 method for class 'LabeledVolumeSet,numeric'
linear_access(x, i)

# S4 method for class 'LatentNeuroVec,numeric'
linear_access(x, i)

# S4 method for class 'LatentNeuroVec,integer'
linear_access(x, i)

Arguments

x

An H5ParcellatedScan object.

i

A numeric vector of 4D linear indices.

...

Additional arguments (not used for this method).

Value

A numeric vector of values corresponding to the provided linear indices. The order of values in the returned vector matches the order of indices in `i`.

A numeric vector of values corresponding to the provided linear indices.

See also

neuroim2::linear_access, specific methods like linear_access,H5ParcellatedScan,numeric-method.

Other H5Cluster: H5ParcellatedMultiScan-class, [,H5ParcellatedScan,ANY,ANY,ANY-method, [,H5ParcellatedScan,ANY,missing,ANY-method, as.data.frame(), as.matrix(), dim(), make_run_full(), make_run_summary(), series(), show,H5ParcellatedScan-method

Other H5Cluster: H5ParcellatedMultiScan-class, [,H5ParcellatedScan,ANY,ANY,ANY-method, [,H5ParcellatedScan,ANY,missing,ANY-method, as.data.frame(), as.matrix(), dim(), make_run_full(), make_run_summary(), series(), show,H5ParcellatedScan-method

Examples

if (FALSE) { # \dontrun{
if (requireNamespace("neuroim2", quietly = TRUE) &&
  requireNamespace("hdf5r", quietly = TRUE) &&
  exists("H5ParcellatedMultiScan", where = "package:fmristore") &&
  exists("linear_access", where = "package:neuroim2") &&
  !is.null(fmristore:::create_minimal_h5_for_H5ParcellatedMultiScan)) {

  temp_exp_file <- NULL
  exp_obj <- NULL
  run_full <- NULL

  tryCatch({
    # Create a minimal H5ParcellatedMultiScan
    temp_exp_file <- fmristore:::create_minimal_h5_for_H5ParcellatedMultiScan(
      master_mask_dims = c(3L, 3L, 2L), # Small dimensions
      num_master_clusters = 2L,
      n_time_run1 = 4L, # For Run1_Full
      n_time_run2 = 0   # No need for Run2_Summary here
    )
    exp_obj <- fmristore::H5ParcellatedMultiScan(file_path = temp_exp_file)

    # Access the H5ParcellatedScan object (helper creates "Run1_Full")
    # The runs() method should give access to the list of runs
    available_runs <- runs(exp_obj)
    run_full <- available_runs[["Run1_Full"]] # Assuming helper creates this scan name

    if (!is.null(run_full)) {
      # Get dimensions: X, Y, Z, T
      run_dims <- dim(run_full) # Should be c(3,3,2,4)
      total_elements <- prod(run_dims)

      # Example: Access first 5 linear indices and last 5
      indices_to_access <- c(1:5, (total_elements - 4):total_elements)
      # Ensure indices are within bounds if total_elements is small
      indices_to_access <- indices_to_access[indices_to_access <= total_elements &
        indices_to_access > 0]
      indices_to_access <- unique(indices_to_access)

      if (length(indices_to_access) > 0) {
        accessed_values <- neuroim2::linear_access(run_full, indices_to_access)
        cat("Accessed values for H5ParcellatedScan:\n")
        print(accessed_values)
        cat("Number of values accessed:", length(accessed_values), "\n")
      } else {
        message("No valid indices to access for linear_access example.")
      }
    } else {
      message("Could not retrieve Run1_Full from the experiment for linear_access example.")
    }

  }, error = function(e) {
    message("linear_access example for H5ParcellatedScan failed: ", e$message)
    if (!is.null(temp_exp_file)) message("Temporary file was: ", temp_exp_file)
  }, finally = {
    if (!is.null(exp_obj)) try(close(exp_obj), silent = TRUE)
    # run_full is part of exp_obj, its resources are managed by exp_obj$close()
    if (!is.null(temp_exp_file) && file.exists(temp_exp_file)) {
      unlink(temp_exp_file)
    }
  })
} else {
  message("Skipping linear_access H5ParcellatedScan example: dependencies/helpers not available.")
}
} # }