v4.0.0

Changelog

v4.0.0

BREAKING CHANGES

All breaking changes are listed here but also kept in their respective sections.

• ⛰️ Features: Add initial support for python 3.12 (#209)
  Support for Python 3.8 is dropped.

⛰️ Features

• [BREAKING CHANGE] Add initial support for python 3.12 (#209)
• Support extras umap with Python 3.12
• Add option to embed data when exporting report to notebook (#206)

🛠️ Refactor

• Use ruff (#217)
• Use mypy type checking (#218)

📚 Documentation

• Fix missing comma in readme (#203)

🧪 Testing

• Rename test modules for consistency (#208)
• Fix imports in multivariate analysis (#211)
• Use fixtures (#207)
• Test data type inference with pyarrow dtypes (#215)

🏗 Build & CI/CD

• Modify publish workflow to allow pypi-trusted-publishing (#202)
• Remove unnecessary publish workflow permissions (#204)
• Update git-cliff (#213)
• Update actions (#210)
• Update poetry to 1.8.2 (#216)
• Fix missing token permissions required for publishing (#219)

⚙️ Miscellaneous Tasks

• Support pandas 2.2 (#205)
• Update pytest (#212)

v3.4.1

Changelog

v3.4.1

🐛 Bug Fixes

• Include readme in pyproject.toml

v3.4.0

Changelog

v3.4.0

⛰️ Features

• Add a warning when a report containing no sections is displayed or exported (#201)

⚙️ Miscellaneous Tasks

• Add support for pandas 2.1 (#199)
• Support seaborn >=0.13 (#200)

v3.3.0

Changelog

v3.3.0

⛰️ Features

• Improve default discrete colorscales in multivariate analysis (#195)

v3.2.2

Changelog

v3.2.2

🐛 Bug Fixes

• Remove deprecated matplotlib parameter (#197)

🛠️ Refactor

• Use enumerate instead of manually incremented index (#194)

🏗 Build & CI/CD

• Bump poetry to version 1.7.1 (#198)

v3.2.1

Changelog

v3.2.1

🐛 Bug Fixes

• Upgrade pyarrow dependency (#193)

🛠️ Refactor

• DRY size_factor calculation (#192)

v3.2.0

Changelog

v3.2.0

⛰️ Features

• Rotate long labels in contingency tables (#188)
• Make contingency table cell size flexible by default (#189)
• Treat categorical color column which is also numeric as categorical in parallel coordinates & parallel categories (#191)

🐛 Bug Fixes

• Fix alignment of colorbar labels for categorical color column in parallel coordinates & parallel categories sections (#190)

📚 Documentation

• Remove author name from changelog (#187)
• Add report demo gif to README (#184)

v3.1.0

Changelog

v3.1.0

⛰️ Features

• Disable missing bar plot by default (#186)

🐛 Bug Fixes

• Data type inference for string types in pandas 2 (#178)
• Suppress warnings in data type inference (#179)
• Move isort from dev dependencies to main dependencies (#181)

🏗 Build & CI/CD

• Fix git-cliff installation (#177)
• Fix typo in publish workflow (#173)
• Execute tests in parallel (#185)

v3.0.0

Changelog

v3.0.0

BREAKING CHANGES

All breaking changes are listed here but also kept in their respective sections.

• 🛠️ Refactor: Use columns arg instead of use_columns and omit_columns (#119)

• 🛠️ Refactor: Add parameter df to method Section.add_cells. (#137)
  Method add_cells of every section has a new required
  parameter df.

• ⛰️ Features: Filter and check columns in section main functions (#133)
  Parameter df removed from constructors of:

  • edvart.report_sections.multivariate_analysis.MultivariateAnalysis,
  • edvart.report_sections.multivariate_analysis.PCA,
  • edvart.report_sections.multivariate_analysis.ParallelCoordinates,
  • edvart.report_sections.multivariate_analysis.ParallelCategories,
  • edvart.report_sections.umap.UMAP.

• 🛠️ Refactor: Remove parameter df from section constructors (#141)
  Parameter df is removed from constructors of
  GroupAnalysis and UnivariateAnalysis.

• 🛠️ Refactor: Move static methods outside their respective classes (#142)
  All report sections class static methods are moved from
  their respective classes to module-level functions. Top-level section
  functions are also renamed to prevent overriding module names (e.g.
  edvart.report_sections.bivariate_analysis.BivariateAnalysis.bivariate_analysis
  -> edvart.report_sections.bivariate_analysis.show_bivariate_analysis).

• 🛠️ Refactor: Make TableOfContents not inherit from Section (#147)

• 🛠️ Refactor: Add prefix show_ to functions matching module name (#155)
  Renamed top-level functions in

• 🛠️ Refactor: Move subsection enums outside section classes (#160)

⛰️ Features

• Treat boolean color column as categorical (#128)
• Treat boolean color col as categorical in scatter plot (#130)
• Exclude swapped column pairs in contingency tables (#131)
• [BREAKING CHANGE] Filter and check columns in section main functions (#133)
• Sort exported notebook imports using isort (#144)
• Add method ReportBase.add_section (#146)
• Add parameter BoxplotsOverTime.group_function_imports (#157)
• Rename inferred data type column (#161)
• Treat date columns with some missing values as dates (#169)
• Add support for pandas 2.0 (#162)

🐛 Bug Fixes

• Fix exported code formatting for UMAP (#139)
• Fix missing functions in exported notebook (#136)
• Fix included columns in contingnecy tables (#145)
• Fix imports in exported code (#153)
• Fix imports when non-default subsections are specified (#151)
• Limit default lags in autocorrelation (#154)
• Imports in time-series rolling statistics (#158)
• Include just year in 'year' column of example dataset (#168)
• Fix default number of histogram bins being extremely high (#167)
• Add missing numpy import in univariate analysis (#175)

🛠️ Refactor

• [BREAKING CHANGE] Use columns arg instead of use_columns and omit_columns (#119)
• Simplify error handling in TimeSeriesAnalysis section (#126)
• Make low verbosity subsection definition more transparent (#127)
• Fix typo in local variable name (#135)
• [BREAKING CHANGE] Add parameter df to method Section.add_cells. (#137)
• [BREAKING CHANGE] Remove parameter df from section constructors (#141)
• [BREAKING CHANGE] Move static methods outside their respective classes (#142)
• [BREAKING CHANGE] Make TableOfContents not inherit from Section (#147)
• Remove quotes around str type hint (#150)
• [BREAKING CHANGE] Add prefix show_ to functions matching module name (#155)
• Move default imports to a constant (#152)
• [BREAKING CHANGE] Move subsection enums outside section classes (#160)
• Fix typos (#170)

📚 Documentation

• Capitalize titles in README.md (#125)
• Add link to docs usage section to README (#134)
• Split examples into basic and TS report (#129)
• Fix link typo (#140)
• Replace bad character (#163)
• Improve contributing guideline (#165)
• Remove module comments (#171)
• Improve report example (#164)

🧪 Testing

• Use non-interactive plotting backend (#138)
• Test exported code executes (#159)

🏗 Build & CI/CD

• Dismiss PR approval on edit (#123)
• Generate release changelog using git-cliff (#143)
• Fix git-cliff installation

⚙️ Miscellaneous Tasks

• Require new umap-learn (#132)
• Remove code for compatibility with pandas <1.4 (#149)
• Add codeowners (#166)
• Support ipywidgets 8+ (#174)

v2.0.0

Changelog

v2.0.0 (2023-09-07)

⚠ BREAKING CHANGE

• The function utils.pair_plot is removed.

• function edvart.utils.is_categorical is removed.edvart.data_types.is_categorical
  can be used instead, with similar behavior.

• Slightly changes behavior of column selection for someplots, e.g. parallel categories is now stricter in choosing which
  columns are considered as categorical.

• Parameter string_representation ofedvart.data_types.infer_data_type removed.
  Call str on the result instead to get the string representation.

• Classedvart.report_sections.dataset_overview.ConstantOccurence renamed to
  ConstantOccurrence.

• Parameter verbosity_constant_occurence ofedvart.report.ReportBase.add_overview renamed to
  verbosity_constant_occurrence.

• Enum value ConstantOccurence ofedvart.report_sections.dataset_overview.Overview.OverviewSubsection
  changed to ConstantOccurrence.

• The name ofedvart.report_sections.timeseries_analysis.TimeAnalysisPlot was
  changed to
  edvart.report_sections.timeseries_analysis.TimeSeriesLinePlot.

• Change Verbosity values. LOW is now 1 (was 0), MEDIUMis 2 (was 1) and HIGH is 3 (was 2).

Features

• remove warning (#104)
  (3c33b6e),
  closes #104
  #94
• Remove utils.is_categorical (#108)
  (39df303),
  closes #108
  #25
• replace warnings with markdown printouts (#94)
  (7ef547b),
  closes #94
  #41
• remove parameter string_representation from infer_data_type (#115)
  (0bab9ca),
  closes #115
• render constant as italic (#114)
  (9fa36a4),
  closes #114
  #57
• Plot pairplot with tight layout. (#91)
  (98b31cf),
  closes #91
• Add UNIQUE data type (#70)
  (c5956ca),
  closes #70
  #33
• change Verbosity enum values (#84)
  (6a2bbcf),
  closes #84
• make addition of ToC idempotent (#74)
  (726f634),
  closes #74
• add equal_scale_axes to 2D scatter plot (#40)
  (307c40a),
  closes #40
• Added dataset_name and dataset_description params to export_notebook (#60)
  (6e411ad),
  closes #60
  #59
• Refactor Report into Report and DefaultReport. (#47)
  (d664e6c),
  closes #47
  #45

Fixes

• Fix usage of missing data type in univariate analysis (#71)
  (7e9cd46),
  closes #71
• set default plotly renderer when exporting to HTML (#99)
  (8c40532),
  closes #99