- table importing checks number of data columns vs. available header entries and raises Warnings/Errors accordingly
- Wavelet analysis parameters automatically change upon changing the sampling interval (sensitive dynamic defaults)
- added
openpyxldependency to re-allow.xlsxdata imports - improved status tips for the settings menu
- Global wavelet spectrum (ensemble average)
- load and save pyBOAT analysis settings to custom .ini file
- redesigned "Analyze All" batch processing menu
- NaN interpolation: now strips of trailing NaNs before interpolation
- repaired NaN interpolation, np.nonzero on boolean Series apparently broke
- minor UI tweaks, further anchor warning boxes
- fixed period axis for single time averaged wavelet spectra
- the UI dataviewer is not plotting envelope + trend anymore, but only the envelope (see here for details)
- plotting the raw signal and detrended signal is mutually exclusive now
- minor UI tweaks, MesageBoxes are anchored to their parent window
- Low-level core.sinc_filter had hardcoded max. size of 2000, removed this such that core.sinc_smooth(..., M=..) can freely select the filter size again. For the GUI the max. size still is 2000, otherwise defaults to signal length if this is smaller.
- hotfix: data loading via import menu repaired
- matplotlib >3.5.1 has a new API for the lines drawn on an axis, made the necessary changes
- improved window closing behavior, 'x' closing the main window should now kill all other open windows
- fixed time-averaged wavelet (Fourier estimate) save routine to also write out the periods
- code formatting via black for all source files
- fixed a bunch of additional UI bugs introduced by newer PyQT versions
- New setting 'Data' for specifying default table output format (csv, txt or xslx)
- just added the link to the gitter chat to the Help menu
- removed bulky talk pdf
- fixed UI bugs concerning ridge smoothing/thresholding and SSG noise strength
- Scripting interface for significance tests with empirical backgrounds, see the new empirical_backgrounds_demo.py
- Added new output settings to control number and graphics formats
- Batch processing allows to export the filtered signals
- New normalization for the average signal power
- New Settings widget to store persistent parameter values
- Reworked the API, see the new
scripting_demo.pyin the repo root - Fixed amplitude envelope plotting issues
- Added a lot more documentation for the API functions
- Fixed one-column txt file parsing error
- capped sinc filter size to 2000 for performance reasons
- No automatic overwrite of analyzer parameters for the synthetic generator
- new taskbar icon ⛵
- Fixed export paths for Windows platforms
- Fixed synthetic signal generator AR(1) alpha setting
- Added batch export of wavelet spectra w/o ridges
- Added time averaging of Wavelet spectra <-> Fourier estimates
- Added Fourier distribution for ensembles, can be used for empirical background estimation
- Reworked FFT visualizations
- Status bars with tool tips for all analysis windows
- Fixed crashes during batch analysis with thresholded ridges
- Warning and on-the-fly interpolation of non-contiguous missing values (NaNs)
- added the pyBOAT icon
- changed version numbering scheme
- fixed Fourier spectra crash
- changed plotting style for short signals
- fixed COI on thresholded ridge
- improved summary statistics
- added batch processing menu
- summary statistics: time-averaged power and ensemble dynamics
- reworked the SSG, catching missing parameter crashes
- Added general data import options with optional missing value interpolation
- Added NaN interpolation to the core module
- Synthetic Signal Generator, yeah!
- Amplitude envelope estimation and removal