ChemAxonAtomFeatureCalculator

Complexity

Total Complexity

26

Size/Duplication

Total Lines	106
Duplicated Lines	29.25 %

Importance

Changes

0

8 Methods

Rating	Name	Duplication	Size	Complexity
A	parse_string()	0	9	4
B	transform()	0	13	6
B	to_padded()	0	16	6
F	_transform_series()	0	33	10
C	__init__()	31	31	7
A	feature_names()	0	3	1
A	_transform_mol()	0	4	1
A	_transform_atom()	0	2	1

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

"""
## skchem.descriptors.atom

Module specifying atom based descriptor generators.
"""

import logging
import subprocess
import tempfile

import pandas as pd

The import pandas could not be resolved.

import numpy as np

The import numpy could not be resolved.

LOGGER = logging.getLogger(__file__)
from ..io import write_sdf, read_sdf

Unused read_sdf imported from io

from .. import core

atom_feats = ['acceptormultiplicity', 'aliphaticatom', 'aromaticatom', 'aromaticelectrophilicityorder',

This line is too long as per the coding-style (103/100).

The name atom_feats does not conform to the constant naming conventions ((([A-Z_][A-Z0-9_]*)|(__.*__))$).

              'asymmetricatom', 'atomicpolarizability', 'chainatom', 'chargedensity', 'chiralcenter', 'distancedegree',

This line is too long as per the coding-style (119/100).

              'donormultiplicity', 'eccentricity', 'electrondensity', 'electrophiliclocalizationenergy', 'hindrance',

This line is too long as per the coding-style (117/100).

              'hmochargedensity', 'hmoelectrondensity', 'hmoelectrophilicityorder', 'hmoelectrophiliclocalizationenergy',

This line is too long as per the coding-style (121/100).

              'hmonucleophilicityorder', 'hmonucleophiliclocalizationenergy', 'ioncharge', 'largestatomringsize', 'nucleophilicityorder',

This line is too long as per the coding-style (137/100).

              'nucleophiliclocalizationenergy', 'oen', 'pichargedensity', 'ringatom', 'ringcountofatom', 'stericeffectindex',

This line is too long as per the coding-style (125/100).

              'totalchargedensity']

# TODO: fix averagemicrospeciescharge

TODO and FIXME comments should generally be avoided.

# TODO: fix logd logp logs

TODO and FIXME comments should generally be avoided.

# TODO: oen (orbital electronegativity) - sigma + pi

TODO and FIXME comments should generally be avoided.

# TODO: water accessible surface area

TODO and FIXME comments should generally be avoided.

class ChemAxonFeatureCalculator(object):

This class should have a docstring.

    _optimal_feats = ['acceptorcount', 'accsitecount', 'aliphaticatomcount', 'aliphaticbondcount', 'aliphaticringcount',

This line is too long as per the coding-style (120/100).

                      'aromaticatomcount', 'aromaticbondcount', 'aromaticringcount', 'asymmetricatomcount',

This line is too long as per the coding-style (107/100).

                      'averagemolecularpolarizability', 'axxpol', 'ayypol', 'azzpol', 'balabanindex', 'bondcount',

This line is too long as per the coding-style (114/100).

                      'carboaliphaticringcount', 'carboaromaticringcount', 'carboringcount', 'chainatomcount',

This line is too long as per the coding-style (110/100).

                      'chainbondcount', 'charge', 'chiralcentercount', 'connectedgraph', 'cyclomaticnumber', 'dipole',

This line is too long as per the coding-style (118/100).

                      'donorcount', 'donorsitecount', 'doublebondstereoisomercount', 'dreidingenergy', 'formalcharge',

This line is too long as per the coding-style (118/100).

                      'fragmentcount', 'fsp3', 'fusedaliphaticringcount', 'fusedaromaticringcount', 'fusedringcount',

This line is too long as per the coding-style (117/100).

                      'hararyindex', 'heteroaliphaticringcount', 'heteroaromaticringcount', 'heteroringcount', 'hlb',

This line is too long as per the coding-style (117/100).

                      'hmopienergy', 'hyperwienerindex', 'largestringsize', 'largestringsystemsize',
                      'markushenumerationcount', 'maximalprojectionarea', 'maximalprojectionradius',
                      'maximalprojectionsize', 'minimalprojectionarea', 'minimalprojectionradius',
                      'minimalprojectionsize', 'mmff94energy', 'molpol', 'pienergy', 'plattindex', 'psa', 'randicindex',

This line is too long as per the coding-style (120/100).

                      'refractivity', 'resonantcount', 'ringatomcount', 'ringbondcount', 'ringcount', 'ringsystemcount',

This line is too long as per the coding-style (120/100).

                      'rotatablebondcount', 'smallestatomringsize', 'smallestringsize', 'smallestringsystemsize',

This line is too long as per the coding-style (113/100).

                      'stereodoublebondcount', 'stereoisomercount', 'szegedindex', 'tetrahedralstereoisomercount',

This line is too long as per the coding-style (114/100).

                      'vdwsa', 'volume', 'wateraccessiblesurfacearea', 'wienerindex', 'wienerpolarity']

This line is too long as per the coding-style (103/100).

    _all_feats = ['acc', 'acceptor', 'acceptorcount', 'acceptormultiplicity', 'acceptorsitecount', 'acceptortable',

This line is too long as per the coding-style (115/100).

                  'accsitecount', 'aliphaticatom', 'aliphaticatomcount', 'aliphaticbondcount', 'aliphaticringcount',

This line is too long as per the coding-style (116/100).

                  'aliphaticringcountofsize', 'aromaticatom', 'aromaticatomcount', 'aromaticbondcount',

This line is too long as per the coding-style (103/100).

                  'aromaticelectrophilicityorder', 'aromaticnucleophilicityorder', 'aromaticringcount',

This line is too long as per the coding-style (103/100).

                  'aromaticringcountofsize', 'asa', 'asymmetricatom', 'asymmetricatomcount', 'asymmetricatoms',

This line is too long as per the coding-style (111/100).

                  'atomicpolarizability', 'atompol', 'averagemicrospeciescharge', 'averagemolecularpolarizability',

This line is too long as per the coding-style (115/100).

                  'averagepol', 'avgpol', 'axxpol', 'ayypol', 'azzpol', 'balabanindex', 'bondcount',
                  'canonicalresonant', 'canonicaltautomer', 'carboaliphaticringcount', 'carboaromaticringcount',

This line is too long as per the coding-style (112/100).

                  'carboringcount', 'chainatom', 'chainatomcount', 'chainbondcount', 'charge', 'chargedensity',

This line is too long as per the coding-style (111/100).

                  'chargedistribution', 'chiralcenter', 'chiralcentercount', 'chiralcenters', 'connectedgraph',

This line is too long as per the coding-style (111/100).

                  'cyclomaticnumber', 'dipole', 'distancedegree', 'don', 'donor', 'donorcount', 'donormultiplicity',

This line is too long as per the coding-style (116/100).

                  'donorsitecount', 'donortable', 'donsitecount', 'doublebondstereoisomercount', 'dreidingenergy',

This line is too long as per the coding-style (114/100).

                  'eccentricity', 'electrondensity', 'electrophilicityorder', 'electrophiliclocalizationenergy',

This line is too long as per the coding-style (112/100).

                  'enumerationcount', 'enumerations', 'formalcharge', 'fragmentcount', 'fsp3',
                  'fusedaliphaticringcount', 'fusedaromaticringcount', 'fusedringcount', 'generictautomer',

This line is too long as per the coding-style (107/100).

                  'hararyindex', 'hasvalidconformer', 'hbda', 'hbonddonoracceptor', 'heteroaliphaticringcount',

This line is too long as per the coding-style (111/100).

                  'heteroaromaticringcount', 'heteroringcount', 'hindrance', 'hlb', 'hmochargedensity',

This line is too long as per the coding-style (103/100).

                  'hmoelectrondensity', 'hmoelectrophilicityorder', 'hmoelectrophiliclocalizationenergy', 'hmohuckel',

This line is too long as per the coding-style (118/100).

                  'hmohuckeleigenvalue', 'hmohuckeleigenvector', 'hmohuckelorbitals', 'hmohuckeltable',

This line is too long as per the coding-style (103/100).

                  'hmolocalizationenergy', 'hmonucleophilicityorder', 'hmonucleophiliclocalizationenergy',

This line is too long as per the coding-style (106/100).

                  'hmopienergy', 'huckel', 'huckeleigenvalue', 'huckeleigenvector', 'huckelorbitals', 'huckeltable',

This line is too long as per the coding-style (116/100).

                  'hyperwienerindex', 'ioncharge', 'isoelectricpoint', 'largestatomringsize', 'largestringsize',

This line is too long as per the coding-style (112/100).

                  'largestringsystemsize', 'localizationenergy', 'logd', 'logdcalculator', 'logp', 'logpcalculator',

This line is too long as per the coding-style (116/100).

                  'logs', 'majormicrospecies', 'majorms', 'majorms2', 'majortautomer', 'markushenumerationcount',

This line is too long as per the coding-style (113/100).

                  'markushenumerations', 'maximalprojectionarea', 'maximalprojectionradius', 'maximalprojectionsize',

This line is too long as per the coding-style (117/100).

                  'minimalprojectionarea', 'minimalprojectionradius', 'minimalprojectionsize', 'mmff94energy',

This line is too long as per the coding-style (110/100).

                  'molecularpolarizability', 'molecularsurfacearea', 'molpol', 'moststabletautomer', 'msa', 'msacc',

This line is too long as per the coding-style (116/100).

                  'msdon', 'name', 'nucleophilicityorder', 'nucleophiliclocalizationenergy', 'oen',
                  'orbitalelectronegativity', 'pi', 'pichargedensity', 'pienergy', 'pka', 'pkacalculator', 'pkat',

This line is too long as per the coding-style (114/100).

                  'plattindex', 'pol', 'polarizability', 'polarsurfacearea', 'psa', 'randicindex',
                  'randommarkushenumerations', 'refractivity', 'resonantcount', 'resonants', 'ringatom',

This line is too long as per the coding-style (104/100).

                  'ringatomcount', 'ringbondcount', 'ringcount', 'ringcountofatom', 'ringcountofsize',

This line is too long as per the coding-style (102/100).

                  'ringsystemcount', 'ringsystemcountofsize', 'rotatablebondcount', 'smallestatomringsize',

This line is too long as per the coding-style (107/100).

                  'smallestringsize', 'smallestringsystemsize', 'stereodoublebondcount', 'stereoisomercount',

This line is too long as per the coding-style (109/100).

                  'stericeffectindex', 'sterichindrance', 'szegedindex', 'tautomercount', 'tautomers',

This line is too long as per the coding-style (102/100).

                  'tetrahedralstereoisomercount', 'tholepolarizability', 'topanal', 'topologyanalysistable',

This line is too long as per the coding-style (108/100).

                  'totalchargedensity', 'tpol', 'tpolarizability', 'vdwsa', 'volume', 'wateraccessiblesurfacearea',

This line is too long as per the coding-style (115/100).

                  'wienerindex', 'wienerpolarity']

    def __init__(self, feat_set='all'):

This code seems to be duplicated in your project.

        if feat_set == 'all':
            self.index = self._all_feats
        elif feat_set == 'optimal':
            self.index = self._optimal_feats
        elif feat_set in self._all_feats:
            self.index = [feat_set]
        elif isinstance(feat_set, (list, tuple)):
            valid = np.array([feat in self._all_feats for feat in feat_set])
            if all(valid):
                self.index = feat_set
            else:
                self.index = feat_set
                raise NotImplementedError('Descriptor \'{}\' not available.'.format(np.array(feat_set)[~valid]))

This line is too long as per the coding-style (112/100).

        else:
            raise NotImplementedError('Feature set {} not available.'.format(feat_set))

    def transform(self, obj):

This method should have a docstring.

        if isinstance(obj, core.Mol):
            return self._transform_series(pd.Series(obj)).iloc[0]
        elif isinstance(obj, pd.Series):
            return self._transform_series(obj)
        elif isinstance(obj, pd.DataFrame):
            return self._transform_series(obj.structure)
        elif isinstance(obj, (tuple, list)):
            return self._transform_series(obj)
        else:
            raise NotImplementedError

    def _transform_series(self, series):

        with tempfile.NamedTemporaryFile(suffix='.sdf') as in_file, tempfile.NamedTemporaryFile() as out_file:

This line is too long as per the coding-style (110/100).

            # write mols to file
            write_sdf(series, in_file.name)
            args = ['cxcalc', in_file.name, '-o', out_file.name] + self.index

            LOGGER.info('Running: ' + ' '.join(args))

            # call command line
            subprocess.call(args)
            try:
                finished = pd.read_table(out_file.name).set_index('id')
            except Exception:

Catching very general exceptions such as Exception is usually not recommended.

                finished = None
        finished.index = series.index
        return finished

class ChemAxonAtomFeatureCalculator(object):

This class should have a docstring.

This abstract class does not seem to be used anywhere.

    _all_feats = ['acceptormultiplicity', 'aliphaticatom', 'aromaticatom', 'aromaticelectrophilicityorder',

This line is too long as per the coding-style (107/100).

                  'asymmetricatom', 'atomicpolarizability', 'chainatom', 'chargedensity', 'chiralcenter',

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

                  'distancedegree', 'donormultiplicity', 'eccentricity', 'electrondensity',
                  'electrophiliclocalizationenergy', 'hindrance', 'hmochargedensity', 'hmoelectrondensity', 'hmoelectrophilicityorder',

This line is too long as per the coding-style (135/100).

                  'hmoelectrophiliclocalizationenergy', 'hmonucleophilicityorder', 'hmonucleophiliclocalizationenergy', 'ioncharge', 'largestatomringsize',

This line is too long as per the coding-style (155/100).

                  'nucleophilicityorder', 'nucleophiliclocalizationenergy', 'oen', 'pichargedensity', 'ringatom', 'ringcountofatom',

This line is too long as per the coding-style (132/100).

                  'stericeffectindex', 'totalchargedensity']

    _h_inc_feats = ['acc', 'atomicpolarizability', 'charge', 'distancedegree', 'don',
       'eccentricity', 'hindrance', 'largestatomringsize', 'oen',

Wrong continued indentation.
'eccentricity', 'hindrance', 'largestatomringsize', 'oen',
^ |

       'ringcountofatom', 'smallestatomringsize', 'stericeffectindex']

Wrong continued indentation.
'ringcountofatom', 'smallestatomringsize', 'stericeffectindex']
^ |

    def __init__(self, feat_set='all', include_hs=False, max_atoms=75):

This code seems to be duplicated in your project.

The argument include_hs seems to be unused.

        """
        Args:
            feat_set (str or list<str>):
                The feature sets to calculate.
                - a single identifier as a `str`
                - a list of identifiers
                - 'h_inc' or those that also calculate for Hs.
                - 'all' for all

            max_atoms:
                - The maximum number of atoms available.
        """
        self.max_atoms = max_atoms

        if feat_set in self._all_feats:
            self.index = [feat_set]
        elif feat_set == 'h_inclusive':
            self.index = self._h_inc_feats
        elif feat_set == 'all':
            self.index = self._all_feats
        elif isinstance(feat_set, (list, tuple)):
            valid = np.array([feat in self._all_feats for feat in feat_set])
            if all(valid):
                self.index = feat_set
            else:
                raise NotImplementedError('Descriptor \'{}\' not available.'.format(np.array(feat_set)[~valid]))

This line is too long as per the coding-style (112/100).

            self.feature_names = feat_set
        else:
            raise NotImplementedError('{} feature set is not available'.format(feat_set))

    @property
    def feature_names(self):

This method should have a docstring.

        return self.index

    def transform(self, obj):

This method should have a docstring.

        if isinstance(obj, core.Atom):
            return self._transform_atom(obj)
        elif isinstance(obj, core.Mol):
            return self._transform_mol(obj)
        elif isinstance(obj, pd.Series):
            return self._transform_series(obj)
        elif isinstance(obj, pd.DataFrame):
            return self._transform_series(obj.structure)
        elif isinstance(obj, (tuple, list)):
            return self._transform_series(obj)
        else:
            raise NotImplementedError

    def _transform_atom(self, atom):
        raise NotImplementedError('Cannot calculate atom wise with Chemaxon')

    def _transform_mol(self, mol):
        ser = pd.Series([mol], name=mol.name)
        res = self._transform_series(ser)
        return res.iloc[0]
        # make into series then use self._transform_mol

    def _transform_series(self, series):

        with tempfile.NamedTemporaryFile(suffix='.sdf') as in_file, tempfile.NamedTemporaryFile() as out_file:

This line is too long as per the coding-style (110/100).

            # write mols to file
            write_sdf(series, in_file.name)
            args = ['cxcalc', in_file.name, '-o', out_file.name] + self.index

            LOGGER.info('Running: ' + ' '.join(args))

            # call command line
            subprocess.call(args)
            finished = pd.read_table(out_file.name).set_index('id')

        def to_padded(s):

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

This function should have a docstring.

            res = np.repeat(np.nan, self.max_atoms)

            def parse_string(s):

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

This function should have a docstring.

                if s == '':
                    return np.nan
                elif s == 'false':
                    return 0
                elif s == 'true':
                    return 1
                else:
                    return float(s)

            ans = np.array([parse_string(i) for i in s.split(';')])
            res[:len(ans)] = ans
            return res
        res = np.array([[to_padded(i) for k, i in val.items()] for idx, val in finished.T.items()])

The variable idx seems to be unused.

The variable k seems to be unused.

        res = pd.Panel(res, items=series.index, major_axis=pd.Index(finished.columns, name='cx_atom_desc'),

This line is too long as per the coding-style (107/100).

                        minor_axis=pd.Index(range(self.max_atoms), name='atom_idx'))

Wrong continued indentation.
minor_axis=pd.Index(range(self.max_atoms), name='atom_idx'))
|^

        return res.swapaxes(1, 2) # to be consistent with AtomFeatureCalculator