to-do (wish)list for the python implementation

[danielemarinazzo]

The current version is functioning and dockerized, thanks to @madhur-tandon .

The matlab version is a few functionalities ahead (apart from the SPM plugin, not applicable to python).

Here the changes implemented in the new version of the matlab toolbox.

https://github.com/compneuro-da/rsHRF/blob/a2eafb68a8c7e308f6776e680c36e857730d60a0/update_log.txt

pasted here with some comment for each of them

------ v2.2, 201909 -----------

1. Change the GUI (add surface analysis pannel and Display)
   Nice to have a GUI also in Python, but low priority

2. add rsHRF_viwer.m for HRF shapes visualization
   Same as before. Nice, but low priority

3. add a m-file (rsHRF_mvgc.m) for multivariate Granger causality analysis.
   This is an important part, but something that comes after the HRF part. Also takes some work (and there could be other toolboxes doing it). Leave aside for now.

4. update HRF basis functions, add Gamma/Fourier basis function (more flexible, and supporting finer temporal grid).
   Important addition, implemented here https://github.com/compneuro-da/rsHRF/blob/master/rsHRF_estimation_temporal_basis.m . To do

5. update (s)FIR model, using AR(k) for autocorrelated noise modeling
   Important addition, implemented here https://github.com/compneuro-da/rsHRF/blob/master/rsHRF_estimation_FIR.m (subfunction glsco, where AR is mentioned, new input parameter para.AR)

6. add a m-file (rsHRF_estimation_impulseest.m, see code for help) for Nonparametric impulse response estimation (not included in rsHRF GUI).
   Also important, here the demo https://github.com/compneuro-da/rsHRF/blob/master/demo_code/demo_rsHRF_impulseest.m) and the code
   https://github.com/compneuro-da/rsHRF/blob/33edcb2ca677ab39518bd90156be1cf047fca932/rsHRF_estimation_impulseest.m

---

------ v2.1, 201908 -----------

1. add suface based analysis module.
   Important to read another type of files (GIFTI). In this case what changes is the read/write part.
   Here some python functions to do so https://github.com/nipy/nibabel/blob/master/nibabel/gifti/gifti.py. They should be included in the read/write part.