DiversityConverter

Complexity

Total Complexity

3

Size/Duplication

Total Lines	27
Duplicated Lines	0 %

Test Coverage

Coverage

25%

Importance

Changes	1
Bugs	0	Features	1

3 Methods

Rating	Name	Duplication	Size	Complexity
A	parse_file()	0	2	1
A	synthetic_targets()	0	2	1
A	__init__()	0	17	1

#! /usr/bin/env python
#
# Copyright (C) 2016 Rich Lewis <rl403@cam.ac.uk>
# License: 3-clause BSD

"""
# skchem.data.coverters.example

Formatter for the example dataset.
"""

import os

import pandas as pd

The import pandas could not be resolved.

import numpy as np

The import numpy could not be resolved.

from .base import Converter, contiguous_order, Feature
from ...pipeline import Pipeline
from ...io import read_sdf
from ...cross_validation import SimThresholdSplit
from ...features import MorganFeaturizer
from ...standardizers import ChemAxonStandardizer

class DiversityConverter(Converter):

    """ Example Converter, using the NCI DTP Diversity Set III.  """

    def __init__(self, directory, output_directory, output_filename='diversity.h5'):

The __init__ method of the super-class Converter is not called.

        output_path = os.path.join(output_directory, output_filename)

        ms = self.parse_file(os.path.join(directory, 'structures.sdf'))

The name ms does not conform to the variable naming conventions ([a-z_][a-z0-9_]{2,30}$).

        y = self.synthetic_targets(ms.index)

The name y does not conform to the variable naming conventions ([a-z_][a-z0-9_]{2,30}$).

        pipeline = Pipeline([ChemAxonStandardizer(keep_failed=True)])

The variable pipeline seems to be unused.

        cv = SimThresholdSplit(min_threshold=0.6, n_jobs=-1).fit(ms)

The name cv does not conform to the variable naming conventions ([a-z_][a-z0-9_]{2,30}$).

        train, valid, test = cv.split((70, 15, 15))
        (ms, y, train, valid, test) = contiguous_order((ms, y, train, valid, test), (train, valid, test))

This line is too long as per the coding-style (105/100).

The name ms does not conform to the variable naming conventions ([a-z_][a-z0-9_]{2,30}$).

The name y does not conform to the variable naming conventions ([a-z_][a-z0-9_]{2,30}$).

        splits = (('train', train), ('valid', valid), ('test', test))

        features = [Feature(fper=MorganFeaturizer(), key='X_morg', axis_names=['batch', 'features'])]

This line is too long as per the coding-style (101/100).

        self.run(ms, y, output_path, splits=splits, features=features)

    def parse_file(self, path):

This method should have a docstring.

This method could be written as a function/class method.

        return read_sdf(path).structure

    def synthetic_targets(self, index):

This method should have a docstring.

This method could be written as a function/class method.

        return pd.Series(np.random.choice([0, 1], size=len(index)), index=index)

Final newline missing