我有一个.csv
文件,以这种方式显示(当我在excel中打开文件时)
a,description,b,c #columns
something,"bla,bla,bla",something,something #elements
something,"bla,bla,bla",something,something #elements
something,"bla,bla,bla",something,something #elements
当我在熊猫中导入.csv时
df = pd.read_csv("a.csv")
描述值不带双引号
a description b c
0 something bla,bla,bla something something
1 something bla,bla,bla something something
2 something bla,bla,bla something something
第一种情况
现在,如果我将df 照原样从熊猫导出到.csv df.to_csv(r"a1.csv", index=False, header=True)
,则双引号仍会按预期出现在.csv文件中(当我在excel中打开时) )
a,description,b,c #columns
something,"bla,bla,bla",something,something #elements
something,"bla,bla,bla",something,something #elements
something,"bla,bla,bla",something,something #elements
第二种情况
但是如果我通过在熊猫中屏蔽来减少数据集
df1 = df.loc[:,("description","c")]
description c
0 bla,bla,bla something
1 bla,bla,bla something
2 bla,bla,bla something
,然后使用同一行代码df1.to_csv(r"a2.csv", index=False, header=True)
将此新数据集df1导出到.csv文件,而无需设置quoting=csv.QUOTE_NONE
,双引号现在在.csv中消失了(当我在excel)
description,c #columns
bla,bla,bla,something #elements
bla,bla,bla,something #elements
bla,bla,bla,something #elements
我想知道为什么会这样吗?
而且我还想知道,如果我将最后一个.csv导入到熊猫error: Expected 2 fields in line x, saw x
中,为什么“描述”列中存在的所有逗号都不会引发错误df2 = pd.read_csv("a2.csv")
。
似乎即使没有显示双引号也可以读取并显示