英文:
pandas read_csv displays quotation marks on the first and last item of every row
问题
我的 .csv 文件看起来像这样：
"col1","col2""1","text1""2","This a "TEXT". However, I cannot parse it."
即它包含在一个值内的逗号和引号。在 read_csv() 函数中使用 sep 参数会在每一行的开头和结尾产生引号：
import pandas as pddf = pd.read_csv('test.csv', sep = '","', engine = 'python')df
结果会是这样：
"col1 col2"0 "1 text1"1 "2 This a "TEXT". However, I cannot parse it.""
我该如何才能正确读取我的文件？
英文:
My .csv file looks like this:<br>
"col1","col2""1","text1""2","This a "TEXT". However, I cannot parse it."
I.e. it contains commas and quotation marks within a value.
Using sep parameter in read_csv() function gives quotation marks in the beginning and ending on each line:
import pandas as pddf = pd.read_csv('test.csv', sep = '","', engine = 'python')df"col1 col2"0 "1 text1"1 "2 This a "TEXT". However, I cannot parse it."
What can I do to read my file correctly?
答案1
得分: 0
问题在于您的CSV中的逗号和引号都没有进行转义。使用","作为分隔符是一个巧妙的解决方法，但它会在开头和结尾保留引号。
df.columns = ['col1', 'col2']df['col1'] = df['col1'].str[1:].astype(int)df['col2'] = df['col2'].str[:-1]col1 col20 1 text11 2 This a "TEXT". However, I cannot parse it.
另一种方法是，如果不是寻找","，而是在引号上使用前瞻和后顾：
df = pd.read_csv('test.csv', sep = r'(?<="),(?=")', engine = 'python')(df.applymap(lambda x: x.strip('"')) # 删除所有值开头和结尾的引号.rename(columns = lambda x: x.strip('"')) # 对列名进行相同处理.assign(col1 = lambda x: x.col1.astype(int)) # 将col1更改为整数列)
英文:
The issue is that none of the commas or quotation marks within your CSV are escaped. Using "," as the delimeter is a smart way around it, but it leaves the quotation marks on the start and end.
df.columns = ['col1', 'col2']df['col1'] = df['col1'].str[1:].astype(int)df['col2'] = df['col2'].str[:-1]col1 col20 1 text11 2 This a "TEXT". However, I cannot parse it.
Here's another way, if instead of looking for ",", you instead had a lookahead and a lookbehind for quotation marks:
df = pd.read_csv('test.csv', sep = r'(?<=\"),(?=\")', engine = 'python')(df.applymap(lambda x: x.strip('"')) # remove quotation marks from the start and end of all values.rename(columns = lambda x: x.strip('"')) # same with column names.assign(col1 = lambda x: x.col1.astype(int)) # change col1 to be a column of ints)
答案2
得分: 0
根据您的有趣想法，您还可以将第一个和最后一个引号作为分隔符，然后删除不需要的列：
data = io.StringIO('''"col1","col2""1","text1""2","This a "TEXT". However, I cannot parse it."''')df = pd.read_csv(data, sep=r'"|,\s*|^"|"$', engine='python').iloc[:, 1:-1]
输出：
col1 col20 1 text11 2 This a "TEXT". However, I cannot parse it.
优点是您可以直接获取正确的数据类型（如果需要）：
df.dtypescol1 int64col2 objectdtype: object
英文:
Building on your interesting idea, you can also add the first and last quotes as separators, then drop the unwanted columns:
data = io.StringIO('''"col1","col2""1","text1""2","This a "TEXT". However, I cannot parse it."''')df = pd.read_csv(data, sep=r'","|^"|"$', engine='python').iloc[:, 1:-1]
Output:
col1 col20 1 text11 2 This a "TEXT". However, I cannot parse it.
The advantage is that you directly get the correct types (if needed):
df.dtypescol1 int64col2 objectdtype: object
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论