raveio 0.9.0
We are close to a major release.
Major changes:
-
Pipeline supports
Python
now! (I think this deserves the first place) -
Referencing electrodes does not require users to run
wavelet
anymore (this one wins the second as it allows for more flexible data processing pipelines) -
Added
PipelineCollections
, allowing users to build and run multiple pipelines -
run_wavelet
will trigger caching existing references -
The pipeline errors is properly handled
-
Pipeline error messages also print out the expressions causing the issues
-
Clears pipeline object directory when related errors are raised; in this case pipeline is rescheduled automatically
-
Pipeline also reports warning messages
-
Allow pipelines to fork
-
Added external data path (
extdata
) to pipeline to store large temporary data -
Added
prepare_subject_phase
andprepare_subject_wavelet
to support loading phase and time-frequency data (besides amplitude) -
Pipeline shared scripts uses script directory as working directory
-
Added
lapply_async
that can run in parallel standalone while is compatible withwith_future_parallel
; this function replacesdipsaus::lapply_async2
-
Added a new subject class
RAVEMetaSubject
for meta-analysis -
Added support for
FreeSurfer
recon-all-clinical.sh
, allowing users to apply super-resolution to their clinicalMRI
before surface reconstruction -
Supports when one
NSx
block data contains multiple recordings (discontinued/paused recording in one block) -
rave_export
exports summary inMatlab
format as well (thanks @Kaitlyn) -
Added
cmd_run_r
to generate reproducible shell scripts that run code -
Added
pipeline$visualize
to view the target relationships and schedule graphs -
Added
pipeline$shared_env
to obtain shared environment
Minor changes:
- Default pipeline repository user name has been changed to
rave-ieeg
- Updated
Github
action configurations - Copy
MRI
to derivative folder when importing images fromNIfTI
format (before imaging segmentation andCT
toMRI
registration) - Updated pipeline template to display the latest pipeline dependency
- Pipeline dependency are installed at the same directory as
ravemanager
- Removed improper
S3
generics invalidate_raw_file_lfp
- Pipeline root searching path can be set with temporary values
- Added
is_on_cran
, allowing certain examples and tests to be suppressed duringCRAN
check, while these code are enabled in the local tests - this helps me to keep examples up-to-date
Bug fixes:
- Fixed template module
UI
having extra padding on bottom - Fixed
Blackrock
NEV-3.0
specification configurations - Fixed errors when subject
set_default
andget_default
raise errors when configuration files are corrupted - Fixed a potential code injection issue
- Fixed parallel code not closing socket connections correctly
raveio 0.1.0
Major changes:
- Added pipeline target format interface; an example for
rave-subject
has been implemented - rewrote
as_rave_subject
such that the subject information will be updated when called - Revised error handlers in pipeline generating engine, so that when errors occur, the enclosing target expressions and back-trace will be printed
- Supported brain viewer to be loaded without electrodes via
include_electrodes
argument - Added
coregistration
viaANTs
- Supports
MRI
registering to templates via non-linear transforms
Minor changes:
- When importing electrodes from other formats, the subject used to be required. However, this is totally unnecessary; removed such requirement
- Fixed a small typo that may cause inconsistent electrode labels (
Nolabel
toNoLabel
)
Bug fixes:
- When executing pipeline shared script, it is possible to leak some variables to the global environment; fixed
- Fixing a bug where shared scripts might not be loaded properly in
pipeline_run