Skip to content

v0.3.0

Choose a tag to compare

@znicholls znicholls released this 06 Jan 14:06
· 122 commits to main since this release

Announcements

  • A couple of key APIs were changed to reflect what they are actually doing, see breaking below
  • We have added a deeper dive into the conversion from discrete to continuous data that will help understand this key part of the package, see improved documentation below

Changelog

⚠️ Breaking Changes

    • Changed the input arguments of [discrete_to_continuous][continuous_timeseries.discrete_to_continuous.discrete_to_continuous].
      We have updated so that, rather than taking in a discrete argument,
      we take in an x and a y array and a name.
      This API better represents the separation between
      discrete representations, continuous representations
      and the conversion in between them (which is a different thing again).
      All other discrete to continuous conversion functions were updated to match this change in API.
    • Changed the input arguments of [Timeseries.from_arrays][continuous_timeseries.Timeseries.from_arrays].
      We have updated so that time_axis_bounds is now x and values_at_bounds is now y.
      This update reflects the fact that, depending on the interpolation choice,
      the passed in values will not always end up being the values at the bounds.

    (#19)

🎉 Improvements

  • Added a check to [TimeseriesDiscrete.to_continuous_timeseries][continuous_timeseries.TimeseriesDiscrete.to_continuous_timeseries]
    so that the user is aware if the chosen interpolation choice means that the instance's
    values at bounds are not actually respected.
    The warning can be controlled with the new warn_if_output_values_at_bounds_could_confuse and check_change_func arguments
    to [TimeseriesDiscrete.to_continuous_timeseries][continuous_timeseries.TimeseriesDiscrete.to_continuous_timeseries]. (#19)

📚 Improved Documentation

Changes