1.3.2

New method in DataFrame: fast_add_rows() for quickly adding large datasets to an existing frame.

1.3.1

• PHP 8.4 compliance.

1.3.0

• Further type hinting.
• Compliance with PHPStan code analysis.
• New rolling window methods added to Vector, DataFrame, PackedSequence and PackedArray: Continually apply a callback to a moving fixed window on the dataset.
• Bug fixes.

1.2.0

• Preliminary compatibility with PHP 8.2
• CSVImporter class, that provides an object orientated approach to working with CSV files.
• Importer class now uses CSVImporter as it's underlying engine for all of its CSV methods.
• DataFrame round() now converts datapoints to a string in order to maintain the specified precision.
• Increased type hints
• Now requires Plotlib 1.1.1 or later.

1.1.0

• Dropped compatibility with PHP 7.3 and 7.4
• Compatible with PHP 8.1 (as of RC5)
• Upped PHEXT-core minimum version to 1.1
• Upped PHEXT-plotlib minimum version to 1.1
• Improved type hinting

1.0.0

Current state good enough for a 1.0 release. This release is functionally the same as 0.6.4 other than it now requires PlotLib 1.0 or later.

0.6.4

• New replace() method for DataFrame. Replace the entire value set of a column with a new set of values.
• New method splitby() for Vector. Hooks into equivilent array method in core lib.

** Now requires PHEXT core 1.0.4 as a minimum.

0.6.3

• Bug fixes.
• Improvements to versatility for SMA and EMA classes (implements ArrayAccess, Countable and Iterable).
• Both SMA and EMA now maintain their history of averages that can be recalled at any time.
• Return types on PackedArray/PackedSequence.
• Ability to run a traversal with DOMScraper using a generator.

0.6.2

• Now requires PlotLib 0.5 or later.
• Provides unit tests for charting methods.
• Charting methods box() and hist() of DataFrame now have standardised GD font being used.
• Fixed incorrect return type against box() method of DataFrame.

Improvements and Maintenance

• Improved examples and documentation.
• DataFrame round() method improved to accept a specific set of columns to operate on.
• DataFrame support for import of vertical datasets where data array is an associative array of column headers and column values.
• Return type added to various methods where the result is constant for DataFrame and Vector.
• Minor improvements to some tests for DataFrame.