metadpy.sdt.roc_auc#

metadpy.sdt.roc_auc(data: None, stimuli: None, responses: None, accuracy: None, confidence: None, nRatings: None, nR_S1: Union[list, ndarray], nR_S2: Union[list, ndarray]) → float[source]#

metadpy.sdt.roc_auc(data: None, stimuli: Union[list, ndarray], responses: Union[list, ndarray], accuracy: Union[list, ndarray], confidence: Union[list, ndarray], nRatings: None, nR_S1: None, nR_S2: None) → float

metadpy.sdt.roc_auc(data: DataFrame, stimuli: str, responses: str, accuracy: str, confidence: str, nRatings: int, nR_S1: None, nR_S2: None) → float

Calculate the area under the type 2 ROC curve given from confidence ratings.

Parameters

data: Dataframe containing one “stimuli” and one “response” column. Different column names can also be provided using the stimuli and responses parameters.
stimuli: If a string is provided, should be the name of the column used as stimuli. If a list or an array is provided, should contain the boolean vectors for stimuli. If None (default) and data is a pandas.DataFrame, will be set to Stimuli by default.
responses: If a string is provided, should be the name of the column used as responses. If a list or an array is provided, should contain the boolean vector for responses. If None (default) and data is a pandas.DataFrame, will be set to Responses by default.
accuracy: If a string is provided, should be the name of the column used as accuracy. If a list or an array is provided, should contain the boolean vector for accuracy. If None (default) and data is a pandas.DataFrame, will be set to Accuracy by default. This parameter is optional if stimuli and responses are known.
confidence: If a string is provided, should be the name of the column used as confidence. If a list or an array is provided, should contain the confidence ratings, matching the number of discret ratings provided in the nRatings parameter. If None (default) and data is a pandas.DataFrame, will be set to Confidence by default.
nRatings: Total of available subjective ratings available for the subject. e.g. if subject can rate confidence on a scale of 1-4, then nRatings = 4. Default is None.
nR_S1: Confience ratings (stimuli 1).
nR_S2: Confidence ratings (stimuli 2).

Returns

auc: Area under the type 2 ROC curve.

Raises

ValueError:: If data and both nR_S1 and nR_S2 are missing.

Examples

>>> nR_S1 = [36, 24, 17, 20, 10, 12, 9, 2]
>>> nR_S2 = [1, 4, 10, 11, 19, 18, 28, 39]
>>> roc_auc(nR_S1, nR_S2)
0.6998064266356949