v0.4.0: Merge pull request #25 from KwstasMCPU/develop

[v.0.4.0-beta] - 2023-11-04

Added

• duplication_summary() method has been added. It produces a table with metrics regarding the duplicated records of the given dataset #19.
• duplication_frequency_table() method has been added. It shows how many times a duplicated record appears per the number of its clones #19.

Changed

• renamed the output columns of completeness():
  • 'perc': 'completeness ratio'
  • 'count': 'complete values'
• format argument of completeness() method has been depreciated.
• _conditional_probability() internal class method converted to a standalone internal function.
• Docstring style finalized to numpy.

Fixed

• Fixed MemoryError bug in completeness() caused in cases of datasets with many columns #20.
• Removed unnecessary print statemens from missing_corr_matrix() #21.

v0.3.0

Added

• missing_corr_matrix() method has been added. It calculates the conditional
  probability of a column's value being NULL if another column's value is NULL.

• format argument was added in the completeness_summary() method. If True
  the output is formated with pandas Styler.

Changed

• The outlier detector functions have been renamed to indicate that they
  are intended for internal use. A single leading underscore in front of the functions
  name has been added:

  • outlier_detector_std-> _outlier_detector_std
  • outlier_detector_iqr -> _outlier_detector_iqr
  • outlier_detector_percentiles -> _outlier_detector_percentiles

Fixed

• Fixed bug in outlier_summary() where categorical columns were also present
  in the dataset.

First pre-release

Includes pandas extensions for:

• completeness summary
• outlier summary