- the ASEvaluation class in the weka.attribute_selection module now offers the following methods for attribute transformers like PCA: transformed_header, transformed_data, convert_instance
- ...
- classes.new_instance method can take an options list now as well
- added classes.get_enum method to return the instance of a Java enum item
- added classes.new_instance method to create new instance of Java class
- added typeconv.jstring_list_to_string_list method to convert a java.util.List containing strings into a Python list
- added typeconv.jdouble_to_float method to convert a java.lang.Double to a Python float
- in module typeconv renamed methods: string_array_to_list to jstring_array_to_list, string_list_to_array to string_list_to_jarray, double_matrix_to_ndarray to jdouble_matrix_to_ndarray, enumeration_to_list to jenumeration_to_list, double_to_float to float_to_jfloat
- added weka.timeseries module that wraps the timeseriesForecasting Weka package
- upgraded Weka to 3.9.5
- added weka.core.systeminfo module for obtaining output from weka.core.SystemInfo
- added system_info parameter to weka.core.jvm.start() method
- merged PR #33 (fracpete#33) to better handle associator output
- added AttributeSelectedClassifier meta-classifier to module weka.classifiers
- added AttributeSelection meta-filter to module weka.filters
- added class_index parameter to weka.core.converters.load_any_file and weka.core.converters.Loader.load_file, which allows specifying of index while loading it (first, second, third, last-2, last-1, last or 1-based index).
- added append and clear methods to weka.filters.MultiFilter and weka.classifiers.MultipleClassifiersCombiner to make adding of filters/classifiers easier.
- added attribute_names() method to weka.core.dataset.Instances class
- added subset method to weka.core.dataset.Instances class, which returns a subset of columns and/or rows.
- added method list_property_names to weka.core.classes module to allow listing of Bean property names (which are used by GridSearch and MultiSearch) for a Java object.
- Upgraded Weka to 3.9.4
- added method suggest_package to the weka.core.packages module for suggesting packages for partial class names/package names (NNge or .ft.) or exact class names (weka.classifiers.meta.StackingC)
- the JavaObject.new_instance method now suggests packages (if possible) in case the instantiation fails due to package not installed or JVM not started with package support
- method train_test_split of the weka.dataset.Instances class now creates a copy of itself before applying randomization, to avoid changing the order of data for subsequent calls.
- method create_instances_from_matrices from module weka.core.dataset now works with pure numeric data again
- added sections for creating datasets (manual, lists, matrices) to examples documentation
- added console scripts: pww-associator, pww-attsel, pww-classifier, pww-clusterer, pww-datagenerator, pww-filter
- added serialize, deserialize methods to weka.classifiers.Classifier to simplify loading/saving model
- added serialize, deserialize methods to weka.clusterers.Clusterer to simplify loading/saving model
- added serialize, deserialize methods to weka.filters.Filter to simplify loading/saving filter
- added methods plot_rocs and plot_prcs to weka.plot.classifiers module to plot ROC/PRC curve on same dataset for multiple classifiers
- method plot_classifier_errors of weka.plot.classifiers module now allows plotting predictions of multiple classifiers by providing a dictionary
- method create_instances_from_matrices from module weka.core.dataset now allows string and bytes as well
- method create_instances_from_lists from module weka.core.dataset now allows string and bytes as well
- added wrapper classes for association classes that implement AssociationRuleProducer (package weka.associations): AssociationRules, AssociationRule, item
- added to_source method to weka.classifiers.Classifier and weka.filters.Filter (underlying Java classes must implement the respective Sourcable interface)
- fixed logging setup in weka.core.jvm to avoid global setting global logging setup to DEBUG (thanks to https://github.com/Arnie97)
- upgraded to Weka 3.9.3
- weka.jar now included in PyPi package
- exposed the following methods in weka.classifiers.Evaluation: cumulative_margin_distribution, sf_prior_entropy, sf_scheme_entropy
- upgraded to Weka 3.9.2
- properly initializing package support now, rather than adding package jars to classpath
- added weka.core.ClassHelper Java class for obtaining classes and static fields, as javabridge only uses the system class loader
- added check_for_modified_class_attribute method to FilterClassifier class
- added complete_classname method to weka.core.classes module, which allows completion of partial classnames like .J48 to weka.classifiers.trees.J48 if there is a unique match; JavaObject.new_instance and JavaObject.check_type now make use of this functionality, allowing for instantiations like Classifier(cls=".J48")
- jvm.start(system_cp=True) no longer fails with a KeyError: 'CLASSPATH' if there is no CLASSPATH environment variable defined
- Libraries mtl.jar, core.jar and arpack_combined_all.jar were added as is to the weka.jar in the 3.9.1 release instead of adding their content to it. Repackaged weka.jar to fix this issue (fracpete#5)
- typeconv.double_matrix_to_ndarray no longer assumes a square matrix (fracpete#4)
- len(Instances) now returns the number of rows in the dataset (module weka.core.dataset)
- added method insert_attribute to the Instances class
- added class method create_relational to the Attribute class
- upgraded Weka to 3.9.1
- plot_learning_curve method of module weka.plot.classifiers now accepts a list of test sets; * is index of test set in label template string
- added missing_value() methods to weka.core.dataset module and Instance class
- output variable y for convenience method create_instances_from_lists in module weka.core.dataset is now optional
- added convenience method create_instances_from_matrices to weka.core.dataset module to easily create an Instances object from numpy matrices (x and y)
- initial release of Python3 port