帶有熊貓和多處理的Python IndexError

Question

我正在嘗試將已從squawka.com抓取的許多xml文件轉換為清晰的csv文件。 我找到了一個腳本，據說該腳本可以執行此操作，但是它始終會引發此錯誤：

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "squawka\utils.py", line 227, in export_all_stats
    df = pd.concat(pool.imap(partial_loader, xml_paths), axis=0, ignore_index=True)
  File "C:\Python27\lib\site-packages\pandas\tools\merge.py", line 812, in concat
    copy=copy)
  File "C:\Python27\lib\site-packages\pandas\tools\merge.py", line 842, in __init__
    objs = list(objs)
  File "C:\Python27\lib\multiprocessing\pool.py", line 668, in next
    raise value
IndexError: list index out of range

我正在嘗試使用此腳本中的函數export_all_stats：

import gc
import glob
import logging
import multiprocessing
import numpy as np
import os
import pandas as pd
import re
import warnings
from dateutil import parser
from functools import partial
from lxml import etree


COMPETITIONS = {
    4: 'World Cup',
    5: 'Champions League',
    6: 'Europa League',
    8: 'English Barclays Premier League',
    9: 'Dutch Eredivisie',
    10: 'Football League Championship',
    21: 'Italian Serie A',
    22: 'German Bundesliga',
    23: 'Spanish La Liga',
    24: 'French Ligue 1',
    98: 'US Major League Soccer',
    114: 'Turkish Super Lig',
    129: 'Russian Premier League',
    199: 'Mexican Liga MX - Apertura',
    214: 'Australian A-League',
    363: 'Brazilian Serie A',
    385: 'Mexican Liga MX - Clausura',
}


TIME_SLICE_EVENTS = [
    'action_areas',
    'all_passes',
    'balls_out',
    'blocked_events',
    'cards',
    'clearances',
    'corners',
    'crosses',
    'extra_heat_maps',
    'fouls',
    'goal_keeping',
    'goals_attempts',
    'headed_duals',
    'interceptions',
    'keepersweeper',
    'offside',
    'oneonones',
    'setpieces',
    'tackles',
    'takeons',
]
ALL_STATISTICS = sorted(TIME_SLICE_EVENTS + ['players', 'teams'])

logger = logging.getLogger()
handler = logging.StreamHandler()
formatter = logging.Formatter('%(asctime)s %(name)-12s %(levelname)-8s %(message)s')
handler.setFormatter(formatter)
logger.addHandler(handler)
logger.setLevel(logging.DEBUG)


class SquawkaReport:
    """Squawka match report object.

    :param path: Path to XML-file to generate match report from.
    """

    def __init__(self, path):
        self.__time_slice_events = TIME_SLICE_EVENTS
        self.path = path
        self.xml = self.read_xml(path)

    # See: https://stackoverflow.com/questions/10967551/how-do-i-dynamically-create-properties-in-python
    def __getattr__(self, name):
        if name in self.__time_slice_events:
            return self._parse_timeslice(name)
        else:
            msg = "'{0}' object has no attribute '{1}'"
            raise AttributeError(msg.format(type(self).__name__, name))

    @staticmethod
    def read_xml(path):
        """Read XML file.
        :param path: Path to XML-file.
        :return: XML tree.
        """
        with open(path, 'r') as f:
            data = f.read()
        xml = etree.fromstring(data)
        return xml

    def _parse_timeslice(self, filter_type):
        xpath = '/squawka/data_panel/filters/{filter_type}/time_slice/event'
        return self._get_elements(xpath.format(filter_type=filter_type))

    def _get_elements(self, xpath):
        elements = self.xml.xpath(xpath)
        if elements:
            return self._parse_elements(elements)
        else:
            return None

    def _parse_elements(self, elements):
        parsed = [dict({c.tag: c.text for c in
                        e.getchildren()}.items() + e.attrib.items())
                  for e in elements]
        return parsed

    @property
    def competition(self):
        return re.findall("/(.*)_\d*.xml", self.path)[0]

    @property
    def filters(self):
        filters_element = self.xml.xpath('/squawka/data_panel/filters')
        if filters_element:
            return [ch.tag for ch in filters_element[0].getchildren()]
        # Some match reports don't have data.
        else:
            return None

    @property
    def kickoff(self):
        date = self.xml.xpath("/squawka/data_panel/game/kickoff/text()")[0]
        return parser.parse(date).strftime('%Y-%m-%d %H:%M:%S %z')

    @property
    def match_id(self):
        return int(re.findall("/.*_(\d+).xml", self.path)[0])

    @property
    def name(self):
        return self.xml.xpath("/squawka/data_panel/game/name/text()")[0]

    @property
    def players(self):
        # TODO: Remove non-player elements
        xpath = '/squawka/data_panel/players/player'
        return self._get_elements(xpath)

    @property
    def teams(self):
        xpath = '/squawka/data_panel/game/team'
        return self._get_elements(xpath)

    @property
    def venue(self):
        return self.xml.xpath("/squawka/data_panel/game/venue/text()")[0]

    @property
    def match_info(self):
        info = ({
            'competition': self.competition,
            'kickoff': self.kickoff,
            'match_id': self.match_id,
            'name': self.name,
            'venue': self.venue,
        })
        for team in self.teams:
            for k in ['id', 'short_name']:
                info['_'.join((team['state'], k))] = team[k]
        return info


def stats_from_file(path, statistic, convert=True):
    """Load data for a statistic from file.

    :param path: Path to file.
    :param statistic: Statistic to load (e.g. 'goals_attempts', 'cards').
    :param convert: Process and clean the data (boolean)
    :return pd.DataFrame with data
    """
    report = SquawkaReport(path)
    return stats_from_report(report, statistic, convert)


def stats_from_report(report, statistic, convert=True):
    """Load data for a statistic from a SquawkaReport object.

    :param report: SquawkaReport object
    :param statistic: Statistic to load (e.g. 'goals_attempts', 'cards').
    :param convert: Process and clean the data (boolean)
    :return pd.DataFrame with data
    """
    stats = pd.DataFrame(getattr(report, statistic))
    stats['competition'] = report.competition
    stats['kickoff'] = report.kickoff
    stats['match_id'] = report.match_id
    if convert:
        return convert_export(stats)
    else:
        return stats


def export_all_stats(xml_dir, out_dir, statistics=ALL_STATISTICS, convert=True, n_jobs=None,
                     sequential=('all_passes', 'extra_heat_maps')):
    """Export all statistics from all XML-files in a folder to CSV.

    :param xml_dir: Path to folder containing XML-files
    :param out_dir: Path to folder to save output to
    :param statistics: Statistics to export
    :param convert: Process and clean the data (boolean)
    :param n_jobs: Number of processes to use
    :param sequential: Iterable with statistics to process sequentially (for memory-intensive stats)
    """

    xml_paths = glob.glob(os.path.join(xml_dir, '*.xml'))

    if n_jobs is None:
        n_jobs = multiprocessing.cpu_count() - 1

    if not os.path.exists(out_dir):
        os.makedirs(out_dir)

    pool = multiprocessing.Pool(n_jobs)
    for statistic in statistics:
        if statistic in sequential:
            df = pd.concat((_load_xml(p, statistic) for p in xml_paths), axis=0, ignore_index=True)
        else:
            partial_loader = partial(_load_xml, statistic=statistic)
            df = pd.concat(pool.imap(partial_loader, xml_paths), axis=0, ignore_index=True)
        if convert:
            df = convert_export(df)
        save_path = os.path.join(out_dir, '{}.csv'.format(statistic))
        df.to_csv(save_path, index=False, encoding='utf8')
        logger.debug("Exported %s to %s", statistic, save_path)


def _load_xml(path, statistic):
    """Load XML files ignoring etree.XMLSyntaxErrors.

    :param path: Path to file.
    :param statistic: Statistic to load (e.g. 'goals_attempts', 'cards').
    :return: XML tree (or None on etree.XMLSyntaxError).
    """
    try:
        return stats_from_file(path, statistic)
    except etree.XMLSyntaxError:
        msg = "XML error loading {}, skipping it...".format(path)
        warnings.warn(msg, RuntimeWarning)


def convert_export(df):
    """Convert a statistics export.
    :param df: pd.DataFrame with statistics (see e.g. stats_from_file())
    :return: processed pd.DataFrame
    """

    def parse_indicator(s, indicator):
        return s.notnull() & (s == indicator)  # Nulls are interpreted as False

    convert_cols = {
        'id': 'int',
        'match_id': 'int',
        'mins': 'int',
        'minsec': 'int',
        'secs': 'int',
        'team_id': 'int'
    }
    coordinate_cols = [
        'end',
        'loc',
        'middle',
        'start',
    ]
    indicator_cols = {
        'is_own': 'yes',
        'headed': 'true',  # Note: ignores all falses
        'shot': 'true',  # Note: ignores all falses
    }
    # Convert strings to ints.
    for col in df.columns.intersection(convert_cols):
        df[col] = df[col].replace('', -1)
        df.loc[df[col].isnull(), col] = -1
        df[col] = df[col].astype(convert_cols[col])

    # Convert indicator cols.
    for col in df.columns.intersection(indicator_cols):
        df[col] = parse_indicator(df[col], indicator_cols[col])

    # Convert coordinate cols.
    for col in df.columns.intersection(coordinate_cols):
        df[[col + '_x', col + '_y']] = split_coordinates(df[col])
        df.drop(col, axis=1, inplace=True)

    return df


def split_coordinates(s):
    """Split Series containing strings with coordinates into a DataFrame.

    :param s: pd.Series
    :return: pd.DataFrame with columns 'x' and 'y'
    """
    if s.notnull().all():
        concatenated = s
    else:
        concatenated = s.copy()
        concatenated.loc[concatenated.isnull()] = ','
    split = pd.DataFrame(concatenated.str.split(',').tolist(), columns=['x', 'y'], dtype=float)
    return split.replace('', np.nan)

我對此很陌生，所以我不知道出了什么問題。 我試圖用Google搜索它，但沒有找到答案。 有人可以幫幫我嗎？

Answer 1

我將通過將pool.imap(partial_loader, xml_paths)分配給變量並查看結果的形狀來開始進行故障排除。 將該對象傳遞給pd.concat存在一些問題。 在沒有看到該對象的情況下，我的猜測是該對象實際上沒有要忽略的索引，並且當您設置ignore_index=True它會嘗試查找不存在的內容，因此IndexError：列表索引超出范圍 。