doubt.datasets.fish_toxicity

 1"""Fish toxicity data set.
 2
 3This data set is from the UCI data set archive, with the description being the original
 4description verbatim. Some feature names may have been altered, based on the
 5description.
 6"""
 7
 8import io
 9
10import pandas as pd
11
12from .dataset import BASE_DATASET_DESCRIPTION, BaseDataset
13
14
15class FishToxicity(BaseDataset):
16    __doc__ = f"""
17    This dataset was used to develop quantitative regression QSAR models to predict
18    acute aquatic toxicity towards the fish Pimephales promelas (fathead minnow) on a
19    set of 908 chemicals. LC50 data, which is the concentration that causes death in
20    50% of test fish over a test duration of 96 hours, was used as model response
21
22    {BASE_DATASET_DESCRIPTION}
23
24    Features:
25        CIC0 (float):
26            Information indices
27        SM1_Dz(Z) (float):
28            2D matrix-based descriptors
29        GATS1i (float):
30            2D autocorrelations
31        NdsCH (int)
32            Atom-type counts
33        NdssC (int)
34            Atom-type counts
35        MLOGP (float):
36            Molecular properties
37
38    Targets:
39        LC50 (float):
40            A concentration that causes death in 50% of test fish over a test duration
41            of 96 hours. In -log(mol/L) units.
42
43    Source:
44        https://archive.ics.uci.edu/ml/datasets/QSAR+fish+toxicity
45
46    Examples:
47        Load in the data set::
48
49            >>> dataset = FishToxicity()
50            >>> dataset.shape
51            (908, 7)
52
53        Split the data set into features and targets, as NumPy arrays::
54
55            >>> X, y = dataset.split()
56            >>> X.shape, y.shape
57            ((908, 6), (908,))
58
59        Perform a train/test split, also outputting NumPy arrays::
60
61            >>> train_test_split = dataset.split(test_size=0.2, random_seed=42)
62            >>> X_train, X_test, y_train, y_test = train_test_split
63            >>> X_train.shape, y_train.shape, X_test.shape, y_test.shape
64            ((708, 6), (708,), (200, 6), (200,))
65
66        Output the underlying Pandas DataFrame::
67
68            >>> df = dataset.to_pandas()
69            >>> type(df)
70            <class 'pandas.core.frame.DataFrame'>
71    """
72
73    _url = (
74        "https://archive.ics.uci.edu/ml/machine-learning-databases/"
75        "00504/qsar_fish_toxicity.csv"
76    )
77
78    _features = range(6)
79    _targets = [6]
80
81    def _prep_data(self, data: bytes) -> pd.DataFrame:
82        """Prepare the data set.
83
84        Args:
85            data (bytes): The raw data
86
87        Returns:
88            Pandas dataframe: The prepared data
89        """
90        # Convert the bytes into a file-like object
91        csv_file = io.BytesIO(data)
92
93        # Read the file-like object into a dataframe
94        cols = ["CIC0", "SM1_Dz(Z)", "GATS1i", "NdsCH", "NdssC", "MLOGP", "LC50"]
95        df = pd.read_csv(csv_file, sep=";", header=None, names=cols)
96
97        return df