calculate_gt.m

%% Ground-Truth Power Calculator Workflow
% This script runs the ground-truth workflow for the power calculator. It uses 
% the entire dataset to compute t-test statistics for each functional connectivity 
% (FC) edge or network. A nonzero t-test value indicates the presence of an effect, 
% with the sign of the t-statistic indicating effect direction. Since the full dataset 
% is used, only one repetition is required.
%
% Usage:
%   1. Before running Calculate_gt, open setparams.m and manually update the 
%      parameters to suit your analysis (e.g., dataset file name, directories, etc.).
%   2. Save your changes in setparams.m.
%   3. Run Calculate_gt, which will load the updated parameters, load the dataset, 
%      configure the experiment (with ground_truth enabled), perform the t-tests, 
%      and save the ground-truth effect estimates.
%
% Key Points:
% - Uses the entire dataset (all subjects) for accurate effect size estimation.
% - Only one repetition is performed since a full-sample estimate is computed.
% - The t-test results are used to determine whether each FC edge shows a positive 
%   or negative effect.
% - Ground-truth results are saved for use in downstream power calculations.
%
% Workflow:
%   1. Load or generate the dataset.
%   2. Configure experiment parameters via setparams (ensure you update this file as needed).
%   3. Enable ground_truth mode and disable parallel processing.
%   4. Set up experiment parameters (n_nodes, n_var, n_repetitions) using setup_experiment_data.
%   5. Create the output directory and assign dataset names/atlas.
%   6. Load outcome and brain data from the dataset.
%   7. For each test, infer the test type and compute t-test statistics.
%   8. Set ground-truth-specific parameters and run the benchmarking workflow.
%
% Author: Fabricio Cravo | Date: March 2025

%% This needs to be optimized - Some could easily be moved to rep 
%% There is a lot of unnecesary redundancy added with the activation

addpath('/Users/f.cravogomes/Desktop/Cloned Repos/NBS_Calculator')

% Set working directory to the directory of this script
scriptDir = fileparts(mfilename('fullpath'));
cd(scriptDir);

a = 10;
vars = who;       % Get a list of all variable names in the workspace
vars(strcmp(vars, 'data_matrix')) = [];  % Remove the variable you want to keep from the list
clear(vars{:});   % Clear all other variables
clc;

[current_path, ~, ~] = fileparts(mfilename('fullpath')); % assuming NBS_benchmarking is current folder
addpath(genpath(current_path));

%% Prepare parameters and dataset
Params = setparams();
Params.ground_truth = true;
Params.parallel = false;
Params.subsample_file_type = 'full_file';

rep_cal_function(Params)