Release 2.15.1

• atlas: gene-expression backend and MRI Toronto atlas stretch and squeeze factors @dbirman @oliche #509 #513
• FDR correction (Benjamin-Hochmann) to correct for multiple testing optional @guidomeijer #506
• SpikeSortingLoader can be used with ONE local mode @juhuntenburg #514

Dynamic Pipeline

What's Changed

• Cleanup by @oliche in #501
• Remove hold on tasks if parent is Incomplete by @juhuntenburg in #505
• Widefield2 by @k1o0 in #504
• CI QC Tests by @k1o0 in #507
• Kim atlas by @mayofaulkner in #508
• Task data repo by @mayofaulkner in #510
• Fiberphotometry by @oliche in #512

Full Changelog: 2.14.0...2.15.0

QC changes and SessionLoader

• Adding brainbox.io.one.SessionLoader for standardized loading of session data
• Changes to TaskQC, VideoQC and DLCQC thresholds and aggregation
• Unfreezing numpy version
• Adding probabilistic brain regions lookup in ibllib.atlas.atlas.BrainAtlas.get_labels
• Removing two voxelless areas from beryl atlas
• Minor updates to audio events detection and audio welchogram

2.13.6

• Hotfix: don't overwrite full settings if iblrig on untagged version

2.13.5 Pseudo session creation contrast distributions fixed

Merge pull request #491 from int-brain-lab/hotfix/2.13.5

Hotfix/2.13.5

Hotfix: local paths and version number

Hotfix: datahandler self.local_paths defined in init

Hotfix: one globus

• Hotfix: globus imports were mixed one.globus and one.remote

Release 2.13.0

• Deprecated ibllib.version
• Fix Globus patcher
• Add SpikeSorting Loader samples2times function
• Fix atlas.BrainCoordinate.xyz2i functions to not quietly wrap indices out of volume bounds.
• Set jobs to Held if parent jobs are Started or Abandoned as well
• Reverse matplotlib colorbars in density displays

Video compression jobs and jobs with waiting parents

• Set video compression jobs to priority 90
• Set jobs to Held if parents are Waiting

Large jobs as service

• brainbox.io.one.SpikeSortingLoader: option to load using collection argument
• Restructuring of how jobs are run on local servers, run large jobs as service