spatio_temporal_cluster_1samp : data should be normalized?

💬 Support & Discussions
1

Hi,

I want to use spatio_temporal_cluster_1samp_test(X, n_permutations=n_permutations, threshold=t_threshold, n_jobs=3, tail=1, seed=34). My X data consists of n_subjects, all magnetometer sensors, and 48 epochs. The shape of X1 is (n_subjects, 102, 48). I am comparing this data with the baseline at epoch 0 (i.e., X = X1 - epoch0).

Should I normalize the data by dividing it by the average across all sensors and epochs for each subject?
Best

  • MNE version: e.g. 0.24.0
  • operating system: Windows 10
2

I am not sure, but I would argue that by comparing your signal to a baseline you already normalize the data assuming the baseline is for each epoch?
Also it sounds a bit like this approach would remove evoked data from the epochs, not sure if this is what you want? Are you interested in evoked differences?
Would be interested to hear what the others think about that?

Best,

Carina