REL: v0.2.0

This update adds PyPalettes and better facilitates customisation of the data clock charts. For a look at the available colour combinations, check this tool on the Python Graph Gallery site.

Added

• PyPalettes library dependency, providing 2500+ palettes.
• Colormap reverse flag parameter added to dataclock function.
• Chart polar spine color parameter added to dataclock.
• Chart polar grid color parameter added to dataclock.

Changed

• Wedge label logic moved to dataclocklib.utility.add_wedge_labels.
• Temporal aggregation logic moved to dataclocklib.utility.aggregate_temporal_columns.

REL: v0.1.8

[0.1.8] - 2025-01-20

This release has quite a few minor changes to the function parameters and a large update to the handling of default text annotations. I have also added a basic overview guide to the site documentation, which focuses soley on how the library is used. The old tutorial is still WIP and details using dataclocklib as part of EDA.

Added

• Basic overview guide on documentation site
• Dynamic 'optimal' figure size calculation based on total wedge count.
• Dynamic chart annotation font scaling & spacing adjustment.
• Dynamic polar axis label font scaling based on number of rings.
• Configuration files (dataclocklib/config/) for default chart title & subtitle creation.
• Dataclock kwargs **fg_kw added, which aligns with pyplot.subplots.
  • Figure size (figsize) parameter will be overwritten and must be modified with the returned Figure object.

REL: v0.1.7

[0.1.7] - 2025-01-15

This release has some changes to the function parameters and handling of default text annotations (see documentation). There are also refinements and improvements in the data clock ring and wedge generation logic.

Changed

• Parameter 'default_text' triggers default chart title and subtitle annotations if chart_title & chart_subtitle are None.
• Parameter 'chart_period' for optional annotation below subtitle for dataset reporting period.
• Raises ValueError if data[date_column] Series does not have not a 'datetime64[ns]' dtype.
• Raises ValueError if data is an empty DataFrame.

Fixed

• Ring & wedge value generation inefficiencies (~75% improvement).
• Redundant inner loop for wedge bar creation.
• Divide by zero error when passed an empty DataFrame.
• Leap year ring values changed from 53 to 52 in 'YEAR_WEEK' mode.

REL: v0.1.6

[0.1.6] - 2025-01-14

The focus of this small release was an update to the tutorial, with more in depth analysis of UK Department for Transport traffic accident data. This serves as an example of the use of dataclocklib to identify and visualise temporal patterns in data.

Changed

• Tutorial update.

REL: v0.1.5

[0.1.5] - 2025-01-10

This release marks the beginning of the main functionality of the library, with a basic tutorial Notebook written for the documentation.

Added

• Jupyter Notebook Tutorial for documentation.

Changed

• Sphinx myst-nb extension to allow inclusions of Notebooks as documentation.