阅读MSR将语料库翻译成Pandas

Question

我从MSR下载了MSR释义语料库，并尝试将其加载到dataframe中，但出现以下错误：

import pandas as pd
df = pd.read_csv(r'C:\MSRParaphraseCorpus\msr_paraphrase_test.txt', sep = '\t' )

错误：

^{pr2}$

我看了第34行，它看起来很好。在

  fname = r'C:\MSRParaphraseCorpus\msr_paraphrase_test.txt'
    with open(fname, encoding="utf8") as f:
        content = f.readlines()

    content[34]

输出：

'0\t1268500\t1268733\tAgainst the Japanese currency, the euro was at 135.92/6.04 yen against the late New York level of 136.03/14.\tThe dollar was at 117.85 yen against the Japanese currency, up 0.1 percent.\n'

Answer 1

问题在于第34行的开放式引号（正如我在评论中提到的）。通过传入禁用csv读取器的引用csv.QUOTE_无. 尝试：

import csv
import pandas as pd

df = pd.read_csv(r'C:\MSRParaphraseCorpus\msr_paraphrase_test.txt', sep = '\t', quoting=csv.QUOTE_NONE)