search

xxlya / Fed_ABIDE

impelmentation of https://arxiv.org/pdf/2001.05647.pdf
30 stars 10 forks source link

data preprocess #1

Closed DockerBB closed 3 years ago

DockerBB commented 3 years ago

How to generate the following data? I don't seem to find how to generate these data in the data processing code.

Yufeng98 commented 3 years ago

Hi! Thanks for your message. Sorry about the late reply. Have you run the single.py? in line 41 and 58, it save the split id in ${site}_sub_overlap.h5 files. This step implements a 5-folder cross validation in order to split training and test dataset, and also comfirms that the truncated subject will not fall into both training and test sets. Please feel free to let me if it still does not work.

DockerBB commented 3 years ago

Ok, thank you for the explain. 👍