DNM: Add fALFF, PerAF, and tSNR outputs #1030
Conversation
tsalo
changed the title
Codecov ReportAttention: Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #1030 +/- ##
==========================================
+ Coverage 82.33% 82.43% +0.10%
==========================================
Files 48 48
Lines 5897 5949 +52
Branches 796 802 +6
==========================================
+ Hits 4855 4904 +49
Misses 839 839
- Partials 203 206 +3 ☔ View full report in Codecov by Sentry. |
| write_ndata( | ||
| mean_bold, | ||
| template=self.inputs.preprocessed_bold, | ||
| filename=self._results["mean_bold"], |
There was a problem hiding this comment.
Where does self._results get created?
There was a problem hiding this comment.
It's created automatically by the class, I believe. I define self._results["mean_bold"] in the line before this I think (344).
xcp_d/utils/restingstate.py
Outdated
| # Lomb-Scargle periodogram will have the same scale. | ||
| voxel_data_censored -= np.mean(voxel_data_censored) | ||
| voxel_data_censored /= np.std(voxel_data_censored) | ||
|
|
||
| time_arr = np.arange(0, n_volumes * TR, TR) | ||
| assert sample_mask.size == time_arr.size, f"{sample_mask.size} != {time_arr.size}" | ||
| time_arr = time_arr[sample_mask] |
There was a problem hiding this comment.
0.5 * fs is the nyquist frequency?
xcp_d/utils/restingstate.py
Outdated
|
|
||
| assert alff.size == n_voxels, f"{alff.shape} != {n_voxels}" | ||
|
|
||
| # Add second dimension to array | ||
| # Add second dimension to arrays |
There was a problem hiding this comment.
I like to use .reshape(-1, 1) for padding, but I don't think there's an official best way to do it
There was a problem hiding this comment.
I actually think the Numpy-preferred method might be numpy.expand_dims, but I don't feel strongly about it.
tsalo
changed the title
Add resting-state metric outputs.
Changes proposed in this pull request
uncensored_denoised_bolddata to the ALFF calculation node instead ofinterpolated_filtered_bold, so ALFF and fALFF aren't calculated from bandpass filtered data.