There is a newer version of the record available.

Published March 20, 2020 | Version v0.5.0rc2
Software Open

HBClab/NiBetaSeries: v0.5.0rc2

Authors/Creators

  • 1. Neuroscience Program, University of Iowa
  • 2. Montréal Neurological Institute
  • 3. Florida International University

Contributors

Project leader:

Supervisor:

  • 1. Neuroscience Program, University of Iowa
  • 2. eScience Institute, University of Washington
  • 3. Berkeley Institute for Data Science, University of California Berkeley
  • 4. Social & Affective Neuroscience Laboratory, University of Oregon
  • 5. Hartley Lab, New York University
  • 6. Cognition and Decision Lab, Boston University
  • 7. Integrated Brain Imaging Center, University of Washington
  • 8. Computational Neuroimaging Lab, BioCruces Health Research Institute
  • 9. Montréal Neurological Institute
  • 10. BIL Consultant
  • 11. Research fellow, Turing Institute
  • 12. Florida International University
  • 13. Open source developer/contributor
  • 14. Health, Brain, and Cognition Lab, University of Iowa

Description

Release Notes

This release adds two EXPERIMENTAL flags: --normalize-betas and --no-signal-scaling. I'm especially unsure about the behavior of --normalize-betas in combination with any hrf that contains "derivatives" or "derivatives + dispersion", since I'm unsure when to divide the the beta estimates with the variances of the beta estimates to derive the t-statistic (i.e., normalizing the beta estimates). This release also has given access to residuals from the LSA and LSS interfaces.

Another WARNING: residuals for the LSS interface are not quite as meaningful since I'm averaging the residuals over all the models being generated. This project is not version 1.x.x yet, so be aware aspects of the API may change rapidly.

Thanks to @michaelmack for opening an issue to suggest new features and @PeerHerholz for their contributions!

CHANGES
  • [ENH] add --normalize-betas option (#299) @jdkent
  • [ENH] add --no-signal-scaling option (#298) @jdkent
  • [ENH,FIX] optimally combine temporal + dispersion derivatives for beta estimation (#296) @jdkent
  • [ENH] add residuals to LSS/LSA interfaces (#294) @jdkent
  • [FIX] remove standardize parameter (#292) @jdkent
  • [ENH] Allow niftiimage input (#289) @jdkent
  • fix singularity install instructions in docs (#288) @PeerHerholz

Files

HBClab/NiBetaSeries-v0.5.0rc2.zip

Files (296.5 kB)

Name Size Download all
md5:6536378d768586ab9c54679620c8b317
296.5 kB Preview Download

Additional details

Related works

Is supplement to
https://github.com/HBClab/NiBetaSeries/tree/v0.5.0rc2 (URL)