lore_sa.dataset.TabularDataset
- class lore_sa.dataset.TabularDataset(data: DataFrame, class_name: str = None)[source]
It provides an interface to handle datasets, including some essential information on the structure and semantic of the dataset.
- df
dataframe containing the whole dataset
- Type
pandas.DataFrame
- descriptor
it contains the essential informationregarding each feature. Format:
>>> {'numeric': {'feature name' : { 'index' : <index of feature column>, 'min' : <min value>, 'max' : <max value>, 'mean': <mean value>, 'std': <standard deviation>, 'median': <median value>, 'q1': <first quartile of the distribution>, 'q3': <third quartile of the distribution, }, ..., ..., }, 'categorical: {'feature name': { 'index' : <index of feature column>, 'distinct_values' : <distinct categorical values>, 'value_counts' : {'distinct value' : <elements count>, ... } } }, ... ... ... }
- Type
dict
Methods
__init__
(data[, class_name])from_csv
(filename[, class_name])Read a comma-separated values (csv) file into Dataset object.
from_dict
(data[, class_name])From dicts of Series, arrays, or dicts.
get_categorical_columns
()return the list of values of the target column :return:
get_feature_name
(index)Get the feature name by index :param index: :return: the name of the corresponding feature
get_feature_names
()get_features_names
()get_number_of_features
()get_numeric_columns
()set_class_name
(class_name)Set the class name.
set_descriptor
(descriptor)set_target_label
(descriptor)Set the target column into the dataset descriptor
it creates the dataset descriptor dictionary
- classmethod from_csv(filename: str, class_name: str = None)[source]
Read a comma-separated values (csv) file into Dataset object. :param [str] filename: :param class_name: optional :return:
- classmethod from_dict(data: dict, class_name: str = None)[source]
From dicts of Series, arrays, or dicts. :param [dict] data: :param class_name: optional :return:
- get_feature_name(index)
Get the feature name by index :param index: :return: the name of the corresponding feature
- set_class_name(class_name: str)[source]
Set the class name. Only the column name string :param [str] class_name: :return: