A user-friendly interface to run Task-Informed HATSA (task_hatsa), which incorporates task information during basis shaping and/or alignment refinement.
Usage
task_hatsa(
subject_data_list,
anchor_indices,
spectral_rank_k = 40,
task_data_list = NULL,
task_method = c("lambda_blend", "gev_patch", "core_hatsa"),
opts = task_hatsa_opts(),
verbose = TRUE,
...
)
run_task_hatsa(...)Arguments
- subject_data_list
List where each element is a numeric matrix (`T_i x V_p`) of time-series data for one subject.
- anchor_indices
Integer vector, indices of the canonical anchor parcels (`1` to `V_p`).
- spectral_rank_k
Integer, the desired dimensionality of the primary spectral sketch. Default is 40.
- task_data_list
List (parallel to `subject_data_list`). Each element provides task-related data for a subject. Structure depends on `task_method` and subsequent processing. Default is NULL.
- task_method
Character string: `"lambda_blend"`, `"gev_patch"`, or `"core_hatsa"`. Default: `"lambda_blend"`.
- opts
List of advanced options created by `task_hatsa_opts()`.
- verbose
Logical. Print progress messages? Default `TRUE`.
- ...
Additional arguments passed to `task_hatsa_opts()` if not provided directly via `opts`.