doubt.datasets.protein

  1"""Protein 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 Protein(BaseDataset):
 16    __doc__ = f"""
 17    This is a data set of Physicochemical Properties of Protein Tertiary Structure. The
 18    data set is taken from CASP 5-9. There are 45730 decoys and size varying from 0 to
 19    21 armstrong.
 20
 21    {BASE_DATASET_DESCRIPTION}
 22
 23    Features:
 24        F1 (float):
 25            Total surface area
 26        F2 (float):
 27            Non polar exposed area
 28        F3 (float):
 29            Fractional area of exposed non polar residue
 30        F4 (float):
 31            Fractional area of exposed non polar part of residue
 32        F5 (float):
 33            Molecular mass weighted exposed area
 34        F6 (float):
 35            Average deviation from standard exposed area of residue
 36        F7 (float):
 37            Euclidean distance
 38        F8 (float):
 39            Secondary structure penalty
 40        F9 (float):
 41            Spacial Distribution constraints (N,K Value)
 42
 43    Targets:
 44        RMSD (float):
 45            Size of the residue
 46
 47    Source:
 48        https://archive.ics.uci.edu/ml/datasets/Physicochemical+Properties+of+Protein+Tertiary+Structure
 49
 50    Examples:
 51        Load in the data set::
 52
 53            >>> dataset = Protein()
 54            >>> dataset.shape
 55            (45730, 10)
 56
 57        Split the data set into features and targets, as NumPy arrays::
 58
 59            >>> X, y = dataset.split()
 60            >>> X.shape, y.shape
 61            ((45730, 9), (45730,))
 62
 63        Perform a train/test split, also outputting NumPy arrays::
 64
 65            >>> train_test_split = dataset.split(test_size=0.2, random_seed=42)
 66            >>> X_train, X_test, y_train, y_test = train_test_split
 67            >>> X_train.shape, y_train.shape, X_test.shape, y_test.shape
 68            ((36580, 9), (36580,), (9150, 9), (9150,))
 69
 70        Output the underlying Pandas DataFrame::
 71
 72            >>> df = dataset.to_pandas()
 73            >>> type(df)
 74            <class 'pandas.core.frame.DataFrame'>
 75    """
 76
 77    _url = "https://archive.ics.uci.edu/ml/machine-learning-databases/" "00265/CASP.csv"
 78
 79    _features = range(9)
 80    _targets = [9]
 81
 82    def _prep_data(self, data: bytes) -> pd.DataFrame:
 83        """Prepare the data set.
 84
 85        Args:
 86            data (bytes): The raw data
 87
 88        Returns:
 89            Pandas dataframe: The prepared data
 90        """
 91        # Convert the bytes into a file-like object
 92        csv_file = io.BytesIO(data)
 93
 94        # Load in the dataframe
 95        df = pd.read_csv(csv_file)
 96
 97        # Put the target column at the end
 98        df = df[[f"F{i}" for i in range(1, 10)] + ["RMSD"]]
 99
100        return df