audplot 1.4.6

Changed: split API documentation into sub-pages for each function

Fixed: x-axis scaling for audplot.waveform(). x-axis values now always correspond to the actual number of samples of the signal

Added: support for Python 3.10

Changed: increase speed of audplot.waveform() by factor >100 for long signals

Fixed: require matplotlib!=3.6.1 to avoid failing of audplot.distribution()

Fixed: make audplot.scatter(..., fit=True) plots reproducible by using a fixed seed for bootstrapping

Added: Python 3.9 support

Fixed: contributing text

Removed: Python 3.6 support

Added: fit and order arguments to audplot.scatter()

Changed: audplot.scatter() shows just a scatter plot with default arguments

Changed: show frequency instead of counts on the y-axis for audplot.distribution()

Fixed: audplot.distribution() bins now independently for each distribution, which ensures the same distribution is always plotted the same way

Fixed: building public documentation in CI pipeline

Added: audplot.waveform()

Added: label_aliases argument to audplot.confusion_matrix()

Changed: show grid lines and remove top ad right axes as default

Changed: Use seaborn.histplot() in audplot.distribution()

Fixed: xticks position for time axes

Added: audplot.detection_error_tradeoff()

Added: audplot.human_format()

Added: show_both argument to audplot.confusion_matrix() which allows showing percentage and absolute numbers in the same figure

Changed: switch from True to False as default for percentage argument of audplot.confusion_matrix()

Fixed: install missing libsndfile1 when publishing docs

Fixed: calculate minimum and maximum in scatter() and series()

Added: cepstrum(), signal(), spectrum()

Fixed: URLs to documentation and source code inside Python package

Added: initial release