Version 2.0.0
Version 2.0 of PTSA is a major update which includes several breaking changes
including the removal of several packages and modules, name changes, and
deprecations.
PTSA 2.0 is the last major version to support loading of lab-specific data. In
future releases, all of this functionality will be removed in favor of using
the cmlreaders package which includes the ability to load timeseries data and
transform into a PTSA-compatible format.
Removals
The following packages and modules have been removed:
ptsa.plotting
- see https://github.com/pennmem/ptsa_plot instead (#147)ptsa.stats
- most functionality now exists in SciPy (#147)ptsa.emd
,ptsa.iwasobi
,ptsa.ica
andptsa.wica
(#147)ptsa.filt
- filtering is now contained primarily inptsa.data.filters
(#158)- Old time series class (fb08e6c)
- All modules within
ptsa.data
that worked with the deprecated time series
implementation
Renames
TimeSeriesX
has been renamed toTimeSeries
and is now found in the
ptsa.data.timeseries
module (#141).TimeSeriesX
still exists as an
alias but will give aDeprecationWarning
.- Most modules have been renamed to conform to PEP8 module naming conventions.
These renames are generally not noticed by users if loading readers from the
ptsa.data.readers
namespace.
Other changes
- Reformatted output of
TalReader.read
- The
atlases
field is a nested structured array, rather than an array of dictionaries - Most idioms for getting coordinates and atlas labels for an electrode should continue to work in the new format
- The
channels
field has been changed from an object array of lists to an array of two integers
- The
- Added IPython notebook demonstrating reading localization information with TalReader (thanks @LoganJF!)
- Replaced custom typing system with the
traits
package (b2f4609) - Improved
TimeSeries.to_hdf
(#138)