arthurai.datasets.arthur_example.ArthurExample#

class arthurai.datasets.arthur_example.ArthurExample(name, download_destination_folder=None)#

Bases: object

Class for a user to interface with example data for model analysis and using Arthur

Methods

`get_dataset`	Returns a dataframe of the model's inputs, predictions, ground truth labels, and non input data
`get_ground_truth_data`	Returns the model's ground truth labels from self.dataset
`get_inputs`	Returns the model's input attribute feature values from self.dataset
`get_non_input_data`	Returns the model's non input attributes from self.dataset
`get_predictions`	Returns the model's predicted values from self.predictions

Attributes

`gt_attribute_names`	The list of attribute names which make up the model's ground truth attributes Either GroundTruth or GroundTruthClass
`input_attribute_names`	The list of feature names which make up the model's input attributes
`non_input_attribute_names`	The list of feature names which make up the model's non-input attributes
`pred_attribute_names`	The list of feature names which make up the model's predicted attributes

get_dataset(split=False, test_split_size=None)#

Returns a dataframe of the model’s inputs, predictions, ground truth labels, and non input data

Parameters

split (bool) – bool, whether to return the data split into train/test.
test_split_size (Optional[float]) – the percentage of data to be split into the test set if splitting the data. If None, uses the default test split defined in the example schema.

Return type

Union[DataFrame, Tuple[DataFrame, DataFrame]]

Returns

dataframe of inputs, predictions, ground truth labels, and non input data

get_ground_truth_data(split=False, test_split_size=None, random_state=None)#

Returns the model’s ground truth labels from self.dataset

Parameters

split (bool) – bool, whether to return the data split by reference/inference (AKA train/test), or whether to return a single dataframe
test_split_size (Optional[float]) – the percentage of data to be split into the test set if splitting the data
random_state (Optional[int]) – int, random state for optional split

Return type

Union[DataFrame, Series, Tuple[DataFrame, DataFrame], Tuple[Series, Series]]

Returns

dataframe of model ground truth values

get_inputs(split=False, test_split_size=None, random_state=None)#

Returns the model’s input attribute feature values from self.dataset

Parameters

split (bool) – bool, whether to return the data split by reference/inference (AKA train/test), or whether to return a single dataframe
test_split_size (Optional[float]) – the percentage of data to be split into the test set if splitting the data
random_state (Optional[int]) – int, random state for optional split

Return type

Union[DataFrame, Series, Tuple[DataFrame, DataFrame], Tuple[Series, Series]]

Returns

dataframe of model input features

get_non_input_data(split=False, test_split_size=None, random_state=None)#

Returns the model’s non input attributes from self.dataset

Parameters

split (bool) – bool, whether to return the data split by reference/inference (AKA train/test), or whether to return a single dataframe
test_split_size (Optional[float]) – the percentage of data to be split into the test set if splitting the data
random_state (Optional[int]) – int, random state for optional split

Return type

Union[DataFrame, Series, Tuple[DataFrame, DataFrame], Tuple[Series, Series]]

Returns

dataframe of model non-input attribute data

get_predictions(split=False, test_split_size=None, random_state=None)#

Returns the model’s predicted values from self.predictions

Parameters

split (bool) – bool, whether to return the data split by reference/inference (AKA train/test), or whether to return a single dataframe
test_split_size (Optional[float]) – the percentage of data to be split into the test set if splitting the data
random_state (Optional[int]) – int, random state for optional split

Return type

Union[DataFrame, Series, Tuple[DataFrame, DataFrame], Tuple[Series, Series]]

Returns

dataframe of model prediction data

property gt_attribute_names: List[str]#

The list of attribute names which make up the model’s ground truth attributes Either GroundTruth or GroundTruthClass

Return type: List[str]
Returns: list of ground truth attribute names

property input_attribute_names: List[str]#

The list of feature names which make up the model’s input attributes

Return type: List[str]
Returns: list of input attribute names

property non_input_attribute_names: List[str]#

The list of feature names which make up the model’s non-input attributes

Return type: List[str]
Returns: list of non-input attribute names

property pred_attribute_names: List[str]#

The list of feature names which make up the model’s predicted attributes

Return type: List[str]
Returns: list of predicted attribute names