Releases: kavli-ntnu/opexebo
Development version 0.3.5
Fixes to HDTuning
Adjustments to GridScore to catch exceptions and return NaNs
Development version 0.3.4
Default peak_search
behaviour changed to use default
instead of sep
Raise an Error (not a warning) if sep
is not installed. Recall that as of 0.3.3, sep
will not be installed by default: the user must specify pip install opexebo[full]
Development version 0.3.3
sep
is now an optional dependency instead of required- To install, you must call
pip install opexebo[full]
to get any optional dependencies - If the user attempts to use the
peak_search_sep
but the module is not installed the user will be WARNED, but no errors will be thrown. The code will then fail-over to the alternative routine (which usesskimage
instead)
Development version 0.3.2
Minor fixes to the tuning_curve
function. Where the bin_width is not an exact divisor of 360°, angular_occupancy
and tuning_curve
would yield different bin numbers due to inconsistent calculation. Calculation has been standardised to
bin_number = int(np.ceil(360/bin_width))
In the case of non-exact divisors, this will result in smaller bin sizes than requested.
Development version 0.3.1
Development version 0.3.0
Development version 2.2
Documentation updates
Minor change to tuning_curve_stats to more robust circular shifting (previous implementation ran into issues when shifting from the first or last index)
Development version 2.1
- Bugfixes to most standard functions
- Substantive changes to
analysis.angular_occupancy()
andgeneral.accumulate_spatial()
to handle 1D arrays - Initial work on Sphinx documentation
Development version 2
Version 0.2.0
New functions
- Added Spatial Occupancy, RateMap, RateMapStats+Coherence
- Added GridScoreStats
- Added Angular Occupancy, TuningCurve, TuningCurveStats
- Added Smoothing, EllipseFit
- Added Defaults, a single location for default values
- (Experimental only) Added BorderScore, SpeedScore
- (Experimental only) Added Shuffling
Changed functions:
- Consistent styling: files are labelled in CamelCapsStyle
- All functions are named_in_underscore_style
- Original functions (Autocorrelation, GridScore, PlaceField) all accept Masked Arrays
- General goal is to use maskedArrays rather than NaNs to indicate invalid values.