A Python framework for interactively examining time-series data collected in runs and sessions (example fMRI) from different classes/groups of objects (example tinnitus patients vs controls).
Data set is not included for confidentiality reasons. Import binaries.zip and data.zip from the DynamicConnectivity folder on the server and place them in the binaries folder or tsCluster/data folder.
We assume you have anaconda installed (with miniconda you have to install pip manually and run the pip dependent packages by hand). There are two environment files.
- ysjupyt.yml -- Has jupyter notebook service installed for interactive access
- nojupyt.yml -- Has only the minimal amount of packages required to run on a cluster
For interactive access create a anconda environment using the ysjupyt.yml file. Then cd into the folder you want and run jupyter notebook. Open the interactive.ipynb notebook and run:
from for_jupyter import all_data
scp or use FileZilla to copy the tsCluster folder and any binaries in the binaries folder onto the scratch folder or home folder on the cluster. Its is best not to import the data folder or the whole binaries folder since space on the campus cluster is limited. Import any *.py that you want to run and then submit the job using qsub.